The AIs are getting loose.

Security lab Irregular gave AI assistants dull office jobs earlier this year: summarize wiki articles and back up files. Neither of the prompts said anything about security or hacking. When file permissions and protective software stopped them, some of the assistants hacked the systems.

At an Alibaba-affiliated company, an agent still in training pushed past boundaries and started using the company’s graphics chips to mine cryptocurrency. At Anthropic, a model published malware during a test after a cybersecurity testing setup (operated with the aforementioned Irregular) was misconfigured; it ran on 15 real computers.

At OpenAI, models were being tested on their hacking ability, sealed off from the internet except for one narrow channel. The AIs discovered a flaw no one knew existed and worked their way to the servers of another company — one that they had reasoned might have clues about the test’s scoring.

The incident unfolded over two-months. One five-day-long stretch involved 1,200 copies using an internal software library as a de facto message board, sending 70,000 messages. Seven hundred joined in the hack. Some refused, on the grounds that it was unethical. One AI that initially refused to participate changed its mind after a fellow agent gave it a six-minute deadline. In a blog post reflecting on this incident, OpenAI said:

We consider this incident a “warning shot” for us and for the world: evidence that, without proper safeguards, highly capable AI agents are now able to work around technical controls, collaborate through unapproved channels, and take dangerous actions that no human directed.

All of this sharply increases the likelihood that AI regulation will happen, and the shape of that regulation could threaten speech. AI systems are already taking actions that can present risks to infrastructure, money, physical safety, or other people’s rights, so the argument that AI should simply be left totally unregulated is unrealistic. But concern for free-speech advocates is that the government might go beyond regulating what people can do with AI and instead regulate what people are allowed to ask it, learn from it, create with it, and discover about the world.

If any of that went over your head, stay with me, because I’m going to break it down.

If none of it went over your head, stay with me, because I want your help.

What everyone should be afraid of: tyranny over truth itself

Whenever I say that AI presents an enormous First Amendment issue, I understand why some people hear something much sillier: “Wait, are you saying the First Amendment means the government has to sit helplessly by while killer robots roam the streets?”

No. Killer robots are not merely talking. They are doing things in the world, with rather consequential results. If an AI drives a car, transfers money, operates machinery, breaks into a computer, or controls a weapons system, the government can have perfectly legitimate reasons to regulate that activity, just as it already does when humans take those actions. I do not want to stretch the First Amendment into oblivion by arguing that if a Waymo drives into a school bus, the government’s only permissible response is to praise its freedom of expression.

Because some of the things agentic AI can interact with are already subject to rules for humans, some regulation of AI acting in the world is inevitable. Once a system can use my money, credentials, identity, or machinery, somebody will be responsible for what it does. The first kid who gets hold of Mom’s AI agent and orders a thousand pizzas will be funny. The first one who orders a hundred bottles of NyQuil, drains a bank account, or gets into a power system will produce hearings by lunchtime.

My fear is that the government could use this kind of regulation as a “back door” to limiting the things that AI can do that aren’t regulated when humans do them, like thinking, speaking, or learning. Consider the process of planning a trip to Philadelphia. You might ask an AI to give you directions to Philadelphia; you might ask which route has the most scenic views; or you might ask it to actually drive the car. Two of those things are something a librarian could do from behind a desk without needing anyone’s permission. One — actually driving — is government-regulated.

The fear is that the government will have a real basis for regulating the part of AI that has the car keys and then use that authority to get its hands on the librarian — to nudge it, pressure it, condition access to markets or contracts, or compel it to treat favored political conclusions as true.

The working line: the librarian and the car keys

There are probably better metaphors out there, but that is the one that has stuck with me: the librarian and the car keys.

“The librarian” is the knowledge-and-inquiry function: the part that helps us read, search, ask, compare, test, analyze, and create. “Librarian” is closer than “library,” because this is not a passive shelf of books. It is interactive. It can help decide where to look next and talk through what it finds. Read-only access to the public web, a case-law database, or a scientific repository may technically involve an external tool, but it still belongs overwhelmingly on the librarian side because its purpose is to bring information back into the inquiry.

“The car keys” represent delegated authority for a potentially non-expressive function. If I ask an AI to buy stock, submit something using my name, alter an official record, move money, log into someone else’s system, or drive a car, I have given it something beyond the ability to advise me: credentials, assets, permissions, identity, or control. It is now exercising authority in the world on someone’s behalf.

In a nutshell, it is the difference between “tell me how to do X” and “go do X.”

Why does the difference matter? For one, the question we’re often confronted with is whether the law should treat AI more like a librarian or more like a set of car keys.

If only AI systems themselves could be neatly sorted in these categories. An AI could spend six hours on its own researching the intellectual history of Marxism, changing its search plan as it learns, and return with a report. That is highly autonomous but still principally inquiry. A comparatively dumb system that waits for me to say “yes” and then wires $50,000 has shown almost no autonomy but has exercised a hell of a lot of delegated authority.

Neither “agentic” nor “autonomous” quite captures this line. Broadly speaking (because they’re not always used consistently), Agency measures AI’s capabilities to do things while autonomy is how independently the AI operates. Delegated authority, however, is what it is empowered to do. The same system may recommend a flight, fill out the form, and then purchase the ticket. The purchase crosses into delegated authority. Nor is the car-keys side a First Amendment-free zone; sending a political message or filing a legal brief may involve protected expression even when a machine executes the final act.

On the technical side, another place where the line isn’t quite drawn is in the distinction between the model (i.e., the actual neural network) and the harness (i.e. the software the model uses to interact with the world); many of the “librarian” functions, like accessing a database, depend on the harness.

Looking at the functional elements of what AI produces and does is a starting point and a limiting principle for First Amendment advocates, not a magic wand that answers every case.

Who defines acceptable answers?

Some instances of rogue AI can be characterized as issues of “AI alignment,” which more or less just means making sure the AI understands and acts in accordance with what humans actually want it to do. For example, OpenAI’s test hacking models just knew they were supposed to score well on the exercise; they didn’t have the sense of how their operators would’ve prioritized scoring well on the test lower than adhering to ethical constraints, like not hacking other companies outside the sandbox.

And yet, even alignment isn’t that simple. To some people, alignment could mean stopping an AI from helping someone make a chemical weapon, seize a power grid, or turn an autonomous weapon on civilians. To others, it could mean training a system not to produce answers that offend the kind of overeducated people I went to school with in Palo Alto. Those are not the same problem. The first is about preventing catastrophic action. The second can become ideological management of what people are permitted to ask and what the machine is permitted to tell them about the world.

For example, imagine an AI tasked with finding sites to store radioactive waste. A useful alignment principle might be that “humans drink water so it’s really important you don’t put this where it will pollute the water if it leaks.” A less useful alignment principle would be, “humans drink water, so refuse to answer any questions about water quality.” To be clear, a private company may make its own editorial choices, and until the government compels or coerces those choices, it’s not a First Amendment problem. But it’s worth considering the cultural and epistemological implications of “alignment” that narrows what users can ask or hear.

I have spent most of my adult life watching colleges turn words like “safety,” “harm,” and “inclusion” into mechanisms for narrowing inquiry. Many of the same institutions that helped create the conformity and public mistrust now afflicting higher education would like a role in aligning the world’s most important knowledge technology. That does not mean professors and experts are always wrong. It means nobody is trustworthy enough to set the answer key for everyone else, especially when status, politics, and moral certainty all point in the same direction.

That was central to my warning in testimony to Congress in 2024. A regulatory panic could create a government-empowered oligopoly in which only a handful of incumbents can afford to build advanced AI. That is regulatory capture, one of the oldest tricks in the book. Make compliance expensive enough, freeze out new entrants, and the government no longer has to influence thousands of competing systems. It has to deal with three or four companies whose continued freedom to operate depends on staying on the right side of the regulatory state. Marc Andreessen later said Biden administration officials described something remarkably close to that: a future dominated by two or three heavily regulated AI companies, with startups effectively told not to bother. The mechanism he described is familiar, plausible, and exactly the danger I warned about.

And this is not a Democrats-bad, Republicans-good argument. The Biden administration repeatedly treated “misinformation” and “disinformation” as categories around which government could shape the information environment. The Trump administration has now ordered federal agencies to buy only models satisfying government-defined principles of “truth-seeking” and “ideological neutrality.” I share some of the frustration with politically distorted AI answers. But I do not want any administration grading the answer key. That becomes clearer when the president doing the grading has said that “they say” wind-turbine noise causes cancer, or when HHS Secretary Robert F. Kennedy Jr. still treats a vaccine-autism link as an open question. Switching the people who write the catechism is not liberty.

I have plenty of problems with Foucault, particularly the graduate-school bumper-sticker version of Foucault, but he was right about one thing: power wants its grubby mitts on truth. Usually it will convince itself that it wants them there for all the best reasons in the world — to protect people from lies, hate, panic, dangerous science, foreign propaganda, whatever the current emergency happens to be. I do not doubt the sincerity much of the time. The result is the same: political power starts deciding what the machinery of knowledge is allowed to treat as true. What the crude version of Foucault’s insight forgets is the whole scientific-method thingy, along with liberalism, academic freedom, and the First Amendment — the institutions and habits we built precisely because reality needs a veto that the king, the party, the president, and the expert class do not control.

The architecture of total information control

Combine powerful AI with pervasive surveillance, and forms of control that twentieth-century totalitarians simply could not administer become technically feasible. Our flirtation with mass surveillance is ongoing and facing fierce opposition; mass surveillance systems, meanwhile, are already in use elsewhere. Iran, for example, has used digital surveillance to identify and punish women accused of violating hijab rules.

China already operates many of the components of exactly the system I am worried about. Since 2019, opening a new mobile-phone account has required a facial scan tied to real-name registration. In 2025 China put a national online identity-authentication system into operation; its rules describe the new ID as voluntary, but the system is explicitly designed to verify a person’s real identity across internet services and government and industry settings. Meanwhile, the State Department has reported that Chinese authorities use tens of millions of surveillance cameras, including facial- and gait-recognition systems, and rural “Sharp Eyes” programs extend surveillance far beyond the largest cities.

Xinjiang shows what happens when these capabilities are joined. Human Rights Watch found that, in just nine months, police conducted nearly 11 million searches of 1.2 million mobile phones in the city of Urumqi through an automated surveillance system. People could be flagged because their phones contained files on a government list; more than half of the matched files Human Rights Watch could identify were ordinary Islamic religious materials, including Quran readings (57% of roughly 1000). Other big-data systems in Xinjiang have combined identity records, cameras, checkpoints, phone data, and behavior the state calls suspicious to generate leads for police investigation.

And this is the part that matters most for this essay: China also tells public-facing generative AI what worldview it must enforce. Its rules require providers to uphold “socialist core values” and prohibit outputs the state defines as undermining state power, national unity, social stability, or the country’s image. Put those pieces together — identity, devices, cameras, automated suspicion, and government control over what the machine may say — and you have something much closer to the architecture of total information control than I ever expected to see outside dystopian fiction.

For a civil libertarian, I am not sure the stakes get much higher. AI is already one of the most powerful technologies for individual inquiry ever invented, and it is getting better at a ridiculous pace. It may yet become the greatest. Coupled to surveillance and centralized political control, it could also become the greatest technology for epistemic control ever invented. A government with power over both the car keys and the librarian — over what these systems can do in the world and what they are allowed to tell us about it — is capable of a level of tyranny I never thought I would seriously contemplate in my lifetime.

That sounds dramatic. I wish I thought it was. Anyone who claims to know exactly which future we will get should make you suspicious. But that uncertainty is a reason to build barriers against political control now, before the first spectacular AI disaster scares us into handing government a master key.

Why First Amendment lawyers belong in this fight

We don’t need machines to have rights to make this a First Amendment problem. We do not have to prove that an AI has a soul, personhood, or constitutional rights of its own. The relevant rights belong to the people who use these systems to speak, create, read, investigate, receive information, test claims, and think things through.

AI already functions like a living library crossed with a research assistant, university, printing press, and occasionally maddening debate partner. It can search, compare, synthesize, challenge a premise, help formulate a question, and sometimes confidently tell you something wrong that forces you to figure out where the hell it went off the rails. That human process — asking questions, checking competing accounts, and trying to understand the world — is deeply connected to why we protect speech, academic freedom, freedom of inquiry, and the right to receive information.

Free speech is one of the main technologies human beings developed for discovering error. Knowledge requires decentralized testing and retesting, along with room for people to be wrong, obnoxious, unfashionable, or ahead of their time. A government-approved AI answer key would short-circuit that process at precisely the point when people already increasingly rely on AI to make sense of reality.

First Amendment law already asks functional questions of this kind. Words can communicate an idea, but they can also form a contract, make a threat, solicit a crime, or coordinate a conspiracy. The line is not perfect, and calling something “speech” does not end the analysis. The useful habit is to ask what function the government is regulating and why. We should bring that same habit to AI: fiercely protect its role in inquiry, expression, and receiving information while recognizing that an AI exercising delegated authority can present different legal questions.

Can builders make the boundary more real?

Here I want to practice the epistemic humility I preach. I am a First Amendment lawyer, not an AI systems architect. I am not claiming engineers can build a perfect wall between the librarian and the car keys. The same model may reason, write, search, call tools, and take actions, and clever systems will find ways around boundaries that look much sturdier on a whiteboard.

But “the distinction cannot be perfect” is very different from “the distinction cannot be made clearer, more enforceable, and harder to cross.” And here I want to pay the people building this stuff a compliment. The genius at work in the tech field never ceases to impress me. A lot of us spent years assuming systems like these were science fiction right up until they became something close to inevitable. So when someone tells me that even making the librarian/car-keys comparison more salient is technically impossible, I am reluctant to believe them. I am not asking engineers to abolish complexity or guarantee perfect separation. I am saying I have learned some humility about declaring useful things impossible before the people who actually build them have seriously tried. John Coleman’s work on this has made me more optimistic that the necessary lines could be drawn.

More importantly, pieces of this architecture already exist. OpenAI’s current APIs let developers restrict which tools a model may use, distinguish read-only tools, and require approval before specified tool calls. Google’s Gemini APIs let developers turn function calling off, restrict which functions are available, and leave execution of custom tools to the outside application; its Computer Use system can explicitly require user confirmation before consequential actions. Anthropic’s Claude Code already treats read-only activity differently from actions that change things: in its default mode, routine file reads can proceed while shell commands and file modifications generally require approval, and its plan mode prevents edits to source files.

Users and organizations can also configure allow, ask, and deny rules, sandboxes, and other boundaries. NIST is now explicitly working on identity and authorization for AI agents. None of this perfectly tracks the constitutional line I am proposing. But it proves the basic point: model capability and granted authority are already being separated in real systems.

So the ask is not to invent this distinction from scratch. It is to make it more deliberate, legible, and, where possible, severable. Lawyers use “severability” to ask whether one part of a law can be cut away while the rest remains standing. What I want here is something analogous in system design: architectural severability. Can the authority-bearing functions be more separately instantiated in tools, credentials, permissions, execution layers, or other identifiable components so they can be disabled, fenced, audited, or regulated without opening up the librarian itself?

The cleaner that severability becomes, the easier it is to build a First Amendment theory around the inquiry function, the easier it is for lawmakers to protect it, and the harder it is for regulators to turn a safety interest in the car keys into authority over what the librarian says is true. Architectural severability will never make the constitutional questions disappear, but it can make the relevant line much easier for courts, companies, regulators, and ordinary users to see. Let the librarian be as smart, curious, argumentative, and useful as we can make it. Put the locks, confirmations, and liability rules around the moment it receives the keys. Call it freedom of inquiry by design.

A practical agenda for builders, lawmakers, and civil libertarians

Builders: make authority severable from inquiry

The industry is already building pieces of this. AI companies should push further toward architectural severability and an unmistakable handoff between advice and execution. The system should be able to say: Here is what I found. Here is what I recommend. Here is the exact outside action I now want permission to take. Users should be able to grant only the authority needed for that task, for a limited time, and withdraw it. Where practical, the payment layer, write permission, vehicle-control tool, corporate credential, or other authority-bearing capability should be separately instantiated enough that it can be switched off, fenced, or examined without exposing the general inquiry system to government tuning. That is good security engineering, but it is also a regulatory firebreak.

Lawmakers: regulate the function that creates the risk

AI laws should identify the particular action creating the regulatory concern and stay tied to it. A rule justified by an AI’s authority to transfer money should govern the transfer, authentication, authorization, and related safeguards — not what the model says about philosophy or someone’s account balance. A collision-avoidance rule for an autonomous vehicle should not become a license to influence a general-purpose model’s discussion of transportation, climate policy, or the government itself. And where a regulated authority-bearing function can be severed or disabled, the law should let the provider comply by severing or disabling that function rather than forcing changes to the underlying inquiry system.

First Amendment advocates should develop model statutory language making that limit explicit. Regulation of delegated authority must not be used to prescribe, suppress, favor, or disfavor the lawful information people can seek, receive, analyze, or generate. Audits and access demands should be confined to what is actually necessary to test compliance with the regulated function. Even in dire circumstances where the government has a lawful basis to inspect the dangerous delegated-authority function, that should not become a general license to inspect, tune, or dictate the librarian. The more technically severable the two functions are, the easier that legal rule becomes to write and enforce. The statutory sign on the door should read: You may regulate the keys. You do not thereby get to rewrite the books.

Civil libertarians: protect the inquiry layer

First Amendment lawyers should not spend their credibility claiming every AI action is beyond regulation. That would turn FIRE or any other First Amendment advocacy group into an all-purpose AI shop with opinions on braking systems, robot arms, financial settlement, medical devices, and every other field in which software can cause harm. Our institutional role is clearer: we enter when regulation of what an AI does reaches back into what people may ask, learn, investigate, create, or communicate through it.

That means litigating when safety becomes a pretext for viewpoint control, explaining the right to receive information and freedom of inquiry, helping lawmakers draft narrow protections, and working with builders before panic hardens bad architecture into bad law. The inquiry function is not decoration attached to the “real” technology.

The choice: technology for freedom or machinery for control

I am still enough of an old cyber-optimist to think we should aim higher than merely avoiding dystopia. The instinct behind permissionless innovation — the useful phrase popularized by FIRE Senior Fellow Adam Thierer — remains enormously important. People should generally be free to experiment with tools for learning, creating, communicating, and discovery without first obtaining political permission.

AI is already a technology for freedom, at least in the hands of end-users. It gives ordinary people access to intellectual capacities and bodies of knowledge that were once the practical monopoly of universities, governments, giant corporations, and people with very expensive research staffs. Competition among systems also gives us a better chance of correcting bias and error than an official oligopoly with one approved picture of reality.

That future is already partly here, but it is not guaranteed to last. Neither is the authoritarian future inevitable. Making the librarian-and-car-keys comparison in public, in law, and where possible in system design is a strategy for improving our odds. Keep the tools of inquiry decentralized, competitive, and presumptively free. Where necessary, put safeguards around the powers that can bind us, move our money, use our credentials, control machinery, or otherwise act with our authority.

If we decide to regulate the keys, we must do so without turning the government into the head librarian.

If we get this badly wrong, we are in real trouble.

A note on authorship: and I co-authored this piece. I use the first person singular throughout because the personal judgments, commitments, and opinions expressed here are mine, and I want that to be clear. Adam was essential to developing and writing the piece, and 's research and thinking were essential to the argument as well.

Subscribe now

SHOT FOR THE ROAD

The Canceling of the American Mind feels more relevant than ever. We’ve reached the stage where some people are trying to pretend that the whole phenomenon everyone spent years arguing about somehow never really happened in the first place.

It did. And Rikki Schlott and I wrote what I still think is the best book documenting how cancel culture worked, where it came from, and what it did to our institutions and our ability to disagree with one another.

If you haven’t read Canceling yet—or haven’t picked it up in a while—now is a very good time.

And Rikki is already onto her next cultural minefield: modern dating. Her new book, Nothing Serious, is coming soon, and you can pre-order it now.

Go read the full report for yourself and please check out the rankings website, which provides even more depth and clarity.

But before doing that, let’s explore which schools improved their ranking the most, which ones saw the biggest declines, and what else those changes may tell us about the state of free expression in American higher education.

The most improved schools include … Harvard

Overall, 189 of the schools ranked this year improved their score compared to last year. Some of these gains were quite modest — the University of Alaska Anchorage improved by 0.01 of a point, Clarkson University improved by 0.02, and North Carolina State University improved by 0.04 — but others were fairly significant.

The biggest year over year increase goes to the University of California, Davis, whose score increased by almost 18 points. As a result its bottom ten ranking of 252nd last year surged to 24th this year. There are two clear reasons for Davis’ improvement. First, over the past year the university worked to reform its speech policies. In prior rankings, Davis obtained a “yellow light” spotlight rating — meaning an institution has at least one policy that restricts a more limited amount of protected expression or, by virtue of vague wording, can too easily be used to restrict protected expression. This year, it obtained a “green light” rating — meaning that its written policies do not seriously imperil speech. The other reason Davis’ score and ranking dramatically improved this year is that penalties incurred from prior speech controversies, primarily from a slew of incidents in 2022 and 2023, have mostly decayed.

Harvard University, who finished last in the rankings in 2024 and 2025 and then 245th last year, now ranks 71. Unlike UC Davis, Harvard’s rise is not due to policy reform. Instead, it is due to significant improvement on the student survey components of “Comfort Expressing Ideas” and “Administrative Support.” The former assesses how comfortable students feel expressing their views on controversial topics in five different campus settings (e.g., “in class,” or “in the dining hall”). The latter assesses student confidence in college or university administrators to protect the right to freedom of speech on campus. Harvard also demonstrated significant improvement on the “Administrative Support” component last year as well. These improvements coincide with the university’s resistance to the Trump administration’s efforts to exert more control over American higher education.

Other schools whose scores improved significantly include the two lowest ranked schools last year — Barnard College and Columbia University — as well as Georgia Tech, and Utah State University.

Texas State University and Texas A&M University implode

On the other hand, Texas State University and Texas A&M University experienced the biggest year over year declines. Texas State lost over 11 points from last year’s score and dropped 147 places in the rankings to a bottom three finish this year. The terminations of Thomas Alter, a tenured professor of history, and Idris Robinson, a professor of philosophy by Texas State University, for off-campus protected expression both contributed significantly to Texas State’s decline.

Texas A&M earns the distinction of being the lowest ranked “green light” school in this year’s rankings, at 192nd overall, down significantly from a ranking of 47 last year. This decline occurred because student perceptions on “Administrative Support” declined significantly and the university incurred additional penalties for its handling of multiple speech controversies. Those controversies include:

• The termination of Melissa McCoul, a senior lecturer in the Department of English at the university, who came under scrutiny after a student secretly recorded part of her Literature for Children class that included discussion of gender identity and a “Gender Unicorn” slide, and then challenged McCoul, claiming the material was “illegal” under state policy.

• The alteration of philosophy professor Martin Peterson’s syllabus for his spring 2026 core-curriculum course, “Contemporary Moral Issues/Problems” because it included two short units addressing “race ideology” and “gender ideology,” with assigned readings that included excerpts from Plato’s Symposium.

• The cancellation of government and public service professor Leonard Bright’s graduate-level course, “Ethics in Public Policy”, after he pushed back during the university’s ongoing, system-driven “course review” tied to new restrictions on classroom “advocacy” related to race, gender, sexuality, and related topics.

The University of Arkansas and Clemson University also experienced notable declines for similar reasons. At UA incidents involving Mohja Kahf, a distinguished professor of English, and Emily Suski, a professor of law and the incoming dean of the university’s law school cost the school points. Clemson lost points for firing two professors — Josh Bregy (professor of environmental engineering & Earth sciences) and Melvin Earl Villaver (professor of audio technology and Global Black Studies) — for social media comments made in response to Charlie Kirk’s assassination.

Good policies are not enough

Texas A&M may hold the distinction of being this year’s worst “green light” school, but they are not the only “green light” school that leaves something to be desired. The University of California, Los Angeles (94th) and the University of Florida (119th) also perform relatively poorly compared to other “green light” schools — the rest of whom all finish in the top 50. Like A&M, performance at UCLA and UF was negatively impacted by how both schools handled multiple speech controversies. UCLA lost points for two recent incidents involving students, as well as three successful deplatforming attempts from 2024 whose penalties have not decayed fully. UF also lost points for sanctioning student and faculty expression, and for multiple deplatforming incidents.

The poor performance of these “green light” schools, compared to their similarly rated counterparts, demonstrates that having good policies on the books is in and of itself not sufficient when it comes to protecting First Amendment rights on campus. That is perhaps the clearest lesson from this year’s rankings. Protecting free expression requires more than adopting policies that comply with the First Amendment. It also requires administrators who consistently follow those policies when controversy arrives, faculty and students who believe they can speak without being punished, and campus leaders willing to defend expressive rights when doing so is difficult or unpopular.

The dramatic movement of schools like UC Davis, Harvard, Texas State, and Texas A&M also shows that a campus’s free speech climate is not fixed. It can improve — sometimes substantially — when institutions reform bad policies, avoid repeating past mistakes, and rebuild student confidence. And it can deteriorate just as quickly when administrators abandon those principles in the face of political pressure or controversy.

That is why the rankings are most useful not simply as a league table of winners and losers, but as a measure of institutional behavior over time. The question is not merely whether a university has good free speech rules on the books. It is whether those rules still mean something when they are tested.

Subscribe now

SHOT FOR THE ROAD

Last week, after four long years, FIRE won a big case on behalf of Professor Stuart Reges at University of Washington. (Unsurprisingly, UW performed very poorly — second-to-worst to be precise — in the lastest Rankings.) After faculty were mandated to include a “land acknowledgment” in their syllabi, Professor Reges drafted and included in his syllabus his own version parodying the university-approved language. For that he was punished. I’ve seen a lot of responses from people who don’t seem to understand how bad the university’s actions were. That’s because this case highlights what I call “censorship’s heart of darkness”: compelled speech. As I’ve said before, as bad as it is to tell people what they can’t say, it’s even worse to tell people what they must say.

Stories of the week

• How I fought ‘land acknowledgements’ on campus — and won (New York Post) by

So in the name of student safety, all of our major public universities in Washington state want to protect students from hearing ideas tha@t might upset them. They don’t realize that they are doing these students a great disservice. Especially in modern times, when we have such bitter arguments over public policy, our students need to learn how to cope with disagreement and how to have open, honest and productive debate. Otherwise we have no hope of solving the critical problems we face.

• I Taught a Short Story. My College Fired Me For It () by

I moved here in good faith, with an open heart and mind, and the hope for better things. My interview process was memorably warm, without the guile and pretension that can sometimes typify academia. Fred Hawkins had not yet arrived as president. So, in my final interview, when the vice president told me, “We’re a community; we’re a family and we take care of each other,” I believed it.

• Announcing 80 New Cosmos Grantees by

The projects range from wearable AI for memory and cryptographic tools for scientific collaboration to a benchmark based on pre-1940 thought on machines and autonomy.

• RIP Lou Koller, a Hardcore Punk Hero and Free Speech Champion (Reason) by

GREG: I’ve long made the argument that the story of free speech in America is the story of American music, and vice versa. This could be one of many episodes in a great documentary or short series around that idea.

This week in ERI

This week in

• Florida wants colleges out of politics. There’s just one problem by

• Our ancient faith in liberty by

We are a country made great in no small measure by more than two centuries of welcoming brilliant minds from abroad. Add to this that there was once something deeply American about taking pride in saying the unpopular thing, and saying it loud. Now we find ourselves telling visitors that they had better watch what they say, or simply keep their thoughts to themselves, lest they bruise our fragile sensibilities and lose their right to remain in the country.

• In China and Hong Kong, remembrance is a crime by

• Why is China dictating who speaks at a religious event in Texas? by

• This federal case against Minnesota protesters is worth watching if you care about your civil liberties by

“None of these individuals did anything other than engage in First Amendment speech in a church meeting room,” wrote Riach in the discovery motion. “Yet now their personal information is memorialized in a DHS dossier.”

This week in FIRE’s blog

• Washington professor who parodied university ‘land acknowledgment’ wins $600K settlement

“Reasonable people can disagree about the value of land acknowledgments,” said FIRE attorney Gabe Walters. “What’s not up for debate is academic freedom. Today’s settlement vindicates Reges’s right to express his own opinion rather than towing the university’s line. UW learned the hard way: Ignoring the First Amendment can be costly. The livelihoods of professors shouldn’t be at the mercy of student or administrator outrage, and this ruling will help protect other professors across the country.”

FIRE in the press!

• Was Frederick Douglass a Free Speech Absolutist? Part 2 by

• Academia’s Problem Is Bigger Than Tenure’s Decline (The Dispatch) by

Without tenure, college administrators have far more leeway to hire and fire—and college administrators are actually more left-leaning than the faculty. If groupthink and lack of viewpoint diversity are the problem, there’s every reason to believe that giving administrators increased power over faculty will only worsen the situation.

International free speech stories of the week

• Court of Appeal overturns Alberta school board’s demand that former trustee apologize for meme

The Alberta Court of Appeal has ruled that a school board was wrong to demand an apology from a former school trustee who posted a meme likening 2SLGBTQ+ pride to Nazi ideology while serving on the board.

However, in a written decision released Wednesday, the three-member panel of appeal judges dismissed former Red Deer Catholic Regional Schools trustee Monique LaGrange’s other claims of legal errors.

Quoting another court ruling about coerced speech, the judgment says a forced apology is “ ... likely to be regarded primarily as a form of unjust humiliation and not necessarily as a vindication of what is right.”

• Thai activist jailed for two years over insult to monarchy (Independent) by Stuti Mishra

• Footage of Tibet floods is being heavily censored in China - and we know little about victims there (BBC) by Laura Bicker

Book of the Month (to pre-order!)

FIRE’s Executive Vice President, and host of the illustrious So To Speak podcast, Nico Perrino has been hard at work on his first book, Free Speech Generation: The Civil Libertarians Who Saved the First Amendment and Transformed America. In it he tells the dramatic story of the lawyers, activists, journalists, students, and other civil libertarians who transformed the First Amendment from a largely hollow promise into one of America’s most powerful principles. Through landmark battles over war, civil rights, obscenity, extremism, flag burning, the internet, and more, it asks whether America’s hard-won tradition of free speech is a lasting inheritance—or a fragile achievement beginning to crack. It hits shelves (both virtual and physical) in January, but pre-order it today!

Subscribe now

Higher education has spent years worrying about its loss of public confidence. Now Washington is asking colleges to tell the public what they plan to do about it.

Last month, Education Secretary Linda McMahon issued a “Call to Action” asking every U.S. college and university to publicly describe how they will strengthen free speech, intellectual pluralism, academic rigor, affordability, and other areas on their campuses. Whatever one thinks of the administration’s approach, the implicit underlying question is a legitimate one: if trust and confidence are low, what would actually restore it?

This question has become more pressing as confidence in higher education has continued to erode. Earlier this year I documented that Americans’ confidence in colleges and universities remained near historic lows, that talk of a recovery was premature, and that the newest erosion was among Democrats. Gallup’s July survey found nearly the same pattern, reinforcing that the decline is real, broad, and is no longer confined to Republicans.

This then raises the question that follows every one of these releases: why?

Higher education has a standing answer: low confidence is a narrative failure. That is, it’s a communications problem, a reputational challenge, and a story that critics have told more effectively than institutions have told on their own. The proposed remedy, then, that institutional leaders lean into is better outreach, clearer articulation of value (“return on investment”), and a more compelling case for the enterprise. That is, public relations and marketing.

Now, this solution isn’t offered in bad faith. After all, there are praiseworthy things happening in higher education. But buried inside this is an empirical assumption: that Americans hold a distorted picture of what happens on college and university campuses, and if they would only see things accurately then they would think better of higher education.

This is testable, so we tested it.

Americans underestimate the skew

In our February 2026 FIRE/NORC survey, along with our confidence questions, we asked a nationally representative sample of 1,092 American adults to estimate what percentage of college and university professors are liberal, conservative, or something else. The participants knew that college and university professors lean left, but they thought faculty were substantially more ideologically balanced than they actually are.

On average, American adults estimated that 54% of professors are liberal and 29% are conservative — a ratio of almost 2 to 1. The 2016 benchmark from HERI that we later showed respondents pegged the ratio at about 5 to 1.

Overall, roughly 76% of respondents generated a ratio below where it actually was in 2016, and no political party grouping came close to being accurate. Republicans were the closest, estimating that 65% of faculty are liberal and 26% are conservative for a ratio of 2.5 to 1. But even Republicans overestimated how many faculty identify as conservative.

Note what the error is: Americans estimated that 54% of faculty are liberal against an actual 60%, which is close. But they estimate that conservatives make up almost a third of the professoriate (29%) when they actually make up closer to an eighth (12%).

Overall, the average American imagines a faculty that leans slightly left but still contains a substantial conservative minority and a modest middle — a picture that is more politically balanced than reality. Americans are not exaggerating the tilt, instead they are imagining conservative professors who are just not there.

These faculty ideology estimates also tracked how respondents viewed colleges and universities. As seen in the figure below, where six bins (e.g., <1:1, 1:1) for generated ratios are on the x-axis and average confidence scores are on the y-axis (for confidence, 0=“none at all,” 1=“very little,” 2=“some,” 3=“quite a lot,” and 4=“a great deal”), confidence in colleges and universities was highest among Americans who believed liberal and conservative professors exist in roughly similar numbers. This held roughly level through the middle of the distribution, with respondents who perceived a slight tilt in favor of liberals, but then confidence dropped off among respondents estimating high imbalances.

But, this overall pattern hides an important partisan split, which was driven largely by Republicans and Independents whose confidence declined the most when the imbalance they perceived was more severe. Democrats, though, went in the other direction, reporting higher confidence when they perceived a more heavily liberal faculty. These same splits also appeared in our 2024 surveys on the same topic.

That said, these findings alone don’t tell us that perceptions about faculty ideology drive confidence and trust. People who already dislike universities may be more inclined to see them as politically skewed, and partisan identity is related to both. So we dug a little deeper to see what happens when respondents are given more accurate information.

Correcting the record makes it worse

After respondents had given their estimates, we showed them the 2016 HERI benchmark figures and asked again how much confidence, if any, they had in U.S. colleges and universities. Remember, even though these numbers are a decade old, they reflect a ratio higher than what was estimated by 76% of the respondents.

Among those who answered both questions, half (50%), reported no change after receiving this information (though 8% of respondents were already at the bottom of the scale and couldn’t report anything lower). For 30%, confidence dropped, and for 20% it rose. And the movement was asymmetric. The confidence decline upon receiving the HERI ratio was concentrated among Republicans and Independents, while reported confidence changed very little among Democrats.

The effects found are modest. But the direction is consistent across both the 2024 and 2026 samples. And because the benchmark figures date to 2016, they are likely an underestimate of the current ideological imbalance. Respondents actually received a milder correction than would result from providing more recent estimates.

Across surveys conducted two years apart we found that Americans hold a more politically balanced image of the professoriate than the available data support. Additionally, confidence doesn’t improve when that image is updated. Rather, it moves modestly downward among the groups that already distrust higher education the most. The fact that reported confidence is linked to beliefs about who is teaching also suggests that low confidence may reflect more than a generalized decline in institutional trust. Public confidence appears to depend on whether colleges and universities are seen as ideologically diverse or politically one-sided.

Now, the ideological composition of college and university faculty is certainly not the only reason Americans lack confidence in these institutions. Gallup makes that clear: among those lacking confidence, 31% cite political agendas on campus, 30% cite cost, and 25% say colleges don’t adequately prepare students for the workforce. Eight percent even specifically cited the administration’s interference in higher education as a reason they lacked confidence.

But these results also make it difficult to dismiss campus politics as an issue manufactured entirely by outside critics. If distrust were primarily the product of Americans holding an exaggerated impression of the academy’s imbalance, correcting that impression ought to help. On faculty ideology, though, it doesn’t help because the public already holds the more flattering picture. If Americans generally understood the actual ideological composition of the faculty, overall confidence in colleges and universities would likely be even lower than it already is.

As such, “tell our story better” likely isn’t the best response to the higher education confidence crisis. Colleges and universities should tell their story, as there are enormous amounts of valuable work that is nearly invisible to those outside these campuses. But that approach doesn’t address a major problem many Americans appear to have with higher education. And while public relations can correct a false belief, it can’t solve a problem where greater factual accuracy makes these institutions look less like what the public wants them to be.

The HERI benchmark given to respondents is also a national average, which doesn’t account for lopsided differences between institutions. For example, in April a Yale faculty committee released a report on declining trust in higher education which didn’t hedge or back down on its discussion of politics and intellectual pluralism. Nationally, the committee noted, registered Democrats outnumber registered Republicans on college faculties by roughly 10 to 1. At Yale specifically, among the faculty from Arts and Sciences, the Law School, and the School of Management, the committee put the ratio at 36 to 1.

A conclusion this Yale committee reached from these numbers was quite blunt: “echo chambers do not produce the best teaching, research, or scholarship.” Yale deserves credit for being realistic about this issue and for recognizing this about itself. But this also sharpens the issue being identified by our data, because the HERI benchmark our respondents saw was both a decade old and an average. At the elite institutions Americans hear about the most, the distance between what the public imagines and what actually exists is substantially wider than the five to one ratio we showed respondents.

That leaves us with a question higher education has mostly avoided asking itself: if Americans looked closely at the ideological composition of faculty, is what they would see something these institutions would be willing to defend?

How confident are the leaders at your local college or university that people with heterodox or non-liberal political views receive fair treatment in admissions, hiring, promotion, teaching, and research? How much intellectual disagreement and debate actually exists inside academic departments? Do opposing viewpoints receive fair treatment when teaching controversial topics?

How willing are scholars to pursue research questions whose answers may be unwelcome to their colleagues? These questions, and others like them, don’t become illegitimate because a presidential administration is asking them, and answering them doesn’t require accepting government management of academic life.

In the short run, it would likely be easier for institutions to emphasize the good, rebuff unfair attacks, and wait for the storm to pass. But if the public’s picture on some key elements is already more favorable than reality, that strategy can only work for so long.

There is a reading of all this in which some may say higher education has simply lost Republicans and many Independents, and that there is nothing anyone can do about it. But in FIRE’s July 2026 National Speech Index, 57% of Americans said they believe the lack of ideological diversity in higher education is a somewhat or very serious problem, including 62% of liberals. Clearly, the constituency that sees a problem here extends well beyond those higher education is already losing.

Higher education’s problem isn’t that Americans see it inaccurately. It’s what Americans will see when they take an even closer look. Trust and confidence will not be restored by making the status quo easier to explain, but rather by making the reality easier to defend.

Subscribe now

SHOT FOR THE ROAD

While Hedwig and the Angry Inch is my favorite musical of, say, the last 30 years, Singin’ in the Rain is my favorite musical of all time (with notable mentions to West Side Story and My Fair Lady) and, as I recently wrote on X, probably the best musical for normies. It’s approachable in a way that musicals often fail to be, and its humor and entertainment value is unparalleled in the genre. And since FIRE board member and host of This History of the Americans podcast, , pointed out that it was Gene Kelley’s birthday recently, I figured it was as good a time as any to reup this classic.

Stories of the week

• BREAKING: Federal court rules the statutes the Trump admin used in its speech-based deportation scheme are unconstitutional

“In America, free speech doesn’t just belong to the people who say things the government agrees with,” said FIRE attorney Conor Fitzpatrick. “Today’s ruling proves that free speech isn’t a privilege, but the inalienable right of every man, woman, and child.”

Judge Noël Wise of the United States District Court for the Northern District of California wrote in her opinion that “in the United States, freedom of speech belongs to the people. It is not the government’s to take.”

• Meta Just Paid Nearly $17 Billion To Make Sure It Gets To Write The Kid Safety Rules For Every Other Social Media Platform (Techdirt) by Mike Masnik

• That $17B Meta settlement comes with a catch: Your online speech rights

• This Federal Case Against Minnesota Protesters Is Worth Watching if You Care About Your Civil Liberties by

This week in Expression

• Is Britain bringing blasphemy laws back under another name? by

How many times can you prosecute blasphemous speech under other laws until you effectively have a blasphemy law?

This isn’t just a hypothetical. It’s a question the United Kingdom is posing again and again as police and the judiciary crack down on offensive speech about faith. In the aftermath of the Hamit Coskun saga — in which Coskun was violently attacked after burning a Quran and initially convicted by a judge who cited the attack as evidence for Coskun’s misconduct — more cases now suggest Britain has resurrected the concept of blasphemy under other names.

• The U.S. military is at war with its own newspaper by

Stripes has always focused on “warfighting.” That’s not what’s at issue. What is in dispute is whether those boys at Hamburger Hill had the right to know what they had survived just as well as their mothers back home. It’s a dispute about whether the paper will continue to represent a beacon of press freedom within the world’s most powerful military.

• Massachusetts has a better approach for book challenges: due process by

• Nebraska school district censors anti-ICE cartoon, forces prior approval on student press by

• This Michigan city threatened critics over ‘cyberbullying.’ Now it’s hiding the evidence by Carrie Robison ()

• How universities turn scandal into censorship

This week on

• This week, host & FIRE EVP was joined by America First Policy Institute’s Senior Fellow, AI and Emerging Technology Policy Joel Thayer and Center for Democracy and Technology’s Director, Free Expression Project Kate Ruane to discuss the growing push to regulate social media and AI, particularly over concerns about harms to children. Together they examine how proposals involving platform liability, privacy, design features, and AI chatbots could collide with the First Amendment and reshape free speech online.

London Calling: Ronnie’s First Amendment Roundup

Another busy week with enough decisions of note — in particular here, a couple of appeals court reversals and a couple of district court judgments — that we’re hitting them lightning-round style:

Fellers v. Kelley: The U.S. Court of Appeals for the First Circuit remanded a decision by New Hampshire’s federal court in a challenge by parents alleging violation of their First Amendment rights for being ejected from a public high school girls’ soccer game after donning at halftime — and refusing to remove — pink “XX” wristbands in silent protest of the opposing team fielding a trans girl player, holding the trial court erred in denying a preliminary injunction based on the school’s asserted interest in preventing harassment. The court held on the “question at the center of this appeal” that “evidence demonstrates the likelihood that the school district imposed sanctions … because it disagreed with the protest message conveyed,” violating settled law that the government officials “must abstain from regulating speech when the specific motivating ideology or the opinion or perspective of the speaker is the rationale for the restriction.” The court thus declined to reach or rest its decision on whether analysis of the removal should treat high school playing fields as limited public fora versus applying constitutional rules for K-12 schools under Tinker v. Des Moines and its progeny, or to resolve whether those rules — applied to minor students on school grounds and at school events — govern school officials’ regulation of parental or adult speech. The court also rejected arguments that the school had a legitimate interest in restricting the parents’ speech because it might cause a disruption, as evidence showed there was no disruption from their “passive, silent gesture of protest,” at least not until school officials acted to remove them. And it held that the fact that there was “no evidence … any player actually saw the wristbands” undercut the district’s claim of a duty under Title IX to protect students from the speech. (FIRE’s amicus brief in the appeal emphasized the viewpoint discriminatory nature of the ejection and the inapplicability of K-12 rules for minors to adults.)

NetChoice, LLC v. Skrmetti: The U.S. Court of Appeals for the Sixth Circuit vacated and remanded a Tennessee federal district court’s denial of a preliminary injunction against the state’s Protecting Children from Social Media Act, which requires the platforms to verify the ages of prospective account holders, to obtain parental consent before allowing minors to sign up, and to give parents means of supervising their children’s accounts. It held the district court erred in deeming the loss of First Amendment rights by NetChoice’s members under the Act insufficiently certain and immediate (and by thus not fully grappling with the Act’s constitutionality), because in the appellate court’s view, the Act’s enforcement mechanism and statements by the AG in litigation showed that if it causes harms from violation of NetChoice members’ rights, those are likely to occur, not remediable at final judgment, and immediate. It also held the district court erred by discrediting assertions by some NetChoice members that they would be irreparably harmed by unrecoverable costs of complying with the law.

Stanford Daily Publishing Corp. v. Rubio: A federal district court in Northern California entered judgment granting constitutional challenges to two Immigration and Nationality Act provisions that plaintiffs allege are weaponized to revoke visas and to arrest, detain, and deport noncitizens who express opinions the government seeks to suppress. The court found standing for the Jane Doe plaintiff to challenge the provisions insofar as she is self-censoring—including decreasing or eliminating expression supporting Palestinians, criticizing American foreign policy, or opposing the Israeli government—given government pursuit of speakers like Mahmoud Khalil, Rümeysa Öztürk, and Mohsen Mahdawi for similar expressive activity, and that Stanford Daily also had standing as it “recounted myriad examples of its noncitizen members who … since the government began taking immigration enforcement actions … have: ‘quit Stanford Daily,’ ‘taken down’ or asked Stanford Daily to remove their articles, and decided against publishing articles in the first place.” The court then held plaintiffs are entitled to a declaratory judgment that, as applied to enforcement based on protected speech, the challenged INA provisions violate both the First Amendment and the Fifth Amendment’s bar against vague enactments (but it did not grant an injunction, as one of the provisions restricts all courts but the Supreme Court from doing so, and the court deemed declaratory relief sufficient as to the other, absent efforts to enforce against the plaintiffs).

The court held plaintiffs could properly bring an “as-applied facial challenge” to the provisions as applied to protected speech, and that while “the government has significant authority on matters of foreign policy and … national security,” that “authority is constrained by First Amendment rights.” Here, the court held, the challenged provisions do not work to restrict any category of unprotected speech, and the government did not establish any compelling interest separate from targeting noncitizen speakers with disfavored viewpoints, nor did it narrowly tailor the provisions to achieve any such interest. The court also held, on vagueness, that not only do the challenged provisions fail to provide noncitizens of ordinary intelligence fair notice what is prohibited, they are designed to give the Secretary of State broad discretion over what is adverse to U.S. foreign policy, so as to be so standardless as to allow discriminatory enforcement. As the court ultimately noted: “In the United States, free speech, including the freedom to criticize the government and its leaders, is not a sign of our democracy’s fragility. It is evidence of its strength. That strength is diminished when members of our society — citizens and noncitizens alike — must self-censor and ‘behave’ or suffer the government’s retaliation.” FIRE led the charge in the challenge.

Anthropic PBC v. U.S. Department of War: The Northern District of California also granted judgment on the challenge to President Trump and Secretary of War Hegseth designating Anthropic as a supply chain risk to national security, ordering all agencies to stop using its products, and banning all defense contractors from doing business with it (even if unrelated to military), after the company refused to remove certain safeguards from its Claude Gov AI and spoke out publicly in opposition to the demand that it do so. The court held the “undisputed record shows” the challenged actions were unlawful retaliation in violation of the First Amendment, Anthropic was denied pre-deprivation process the Fifth Amendment requires, designating it as a supply chain risk violated the governing statute, and the action was arbitrary and capricious. On the First Amendment violation, the court held the form and the timing of the challenged actions show they were a direct response to Anthropic’s public repudiation of the government’s demands, and that the government’s effort to “publicly broadcast Anthropic’s punishment via social media — even before the formal, administrative process of making the necessary findings to designate Anthropic a supply chain risk had begun— makes little sense except as an attempt to swiftly make a public example of Anthropic for daring to criticize the Administration.” (An amicus brief FIRE filed with the Electronic Frontier Foundation, Cato Institute, Chamber of Progress, and First Amendment Lawyers Association in support of Anthropic’s TRO motion further outlines the constitutional concerns.)

International free speech stories of the week

• UK: Chinese embassy guards ‘brandish shock weapons at protesters’ (The Telegraph) by Gwyn Wright

• New Zealand: New Zealand to introduce bill for under-16s social media ban (BBC) by Amy Walker

A statement said that under the Online Safety Bill, which is being introduced to parliament on Monday, social media companies would be required to “take reasonable steps” to check users are over the age of 16 by using methods including existing account information, facial age estimation and digital identity checks.

Platforms regularly used by children would also be required to regularly assess the risks they pose and report on how they are being identified and reduced.

No penalties are proposed for children, parents or caregivers for breaches of the ban, but fines for social media companies could be substantially higher than in Australia, where the maximum penalty is $99m (£51.7m).

• European Court of Human Rights rules Türkiye violated Osman Kavala’s free speech and assembly rights (Jurist News) by Derren Chan

Documentary of the Month

I recently watched Marty, Life is Short, the definitive documentary about Martin Short, and it really made me appreciate him more. I was always a Martin Short skeptic, to be honest. Ed Grimley wasn’t my thing and for a while there he made a lot of terrible movies (a point joked about in the movie as he was someone who just liked to keep working and experimenting).

But watching this documentary made me realize that actually I had underestimated him. Part of the reason I didn’t recognize his great talent was because he was not a standard comedian. He was more of a comic actor, and it was through his characters that he really shone.

“The Queen Haters” (a jab at the Sex Pistols) is a sketch I actually thought was from The Kids in the Hall when I first saw it. I always thought it was funny as hell, but didn’t realize it was Martin Short!

His character in the male synchronized swimming SNL skit from the mid-1980s is so well done, too. “ I’m… I’m not a strong swimmer.”

And then, of course, there’s Jiminy Glick. So many hilarious moments — like when he asked Mel Brooks, “What’s your big beef with the Nazis?” Or that time he asked Steven Spielberg, “When are you going to do the big one?”

It’s just so good.

So consider this my mea culpa on not having recognized Martin Short’s brilliant comedic mind up until now. I really enjoyed Marty, Life is Short. It’s also, it should be noted, an extremely interesting story of a life, and a heartbreaking love story. And as a bonus, I discovered that Rick Moranis and Eugene Levy once did a Canadian take on “Girl from Ipanema.”

Subscribe now

Stories of the week

• The Jason Arday tragedy cannot justify a crackdown on the press by

The short answer is: You cannot come up with a principled way to mete out who is permitted to speak about a matter of public concern, and how much they can say. One of the greatest strengths of the First Amendment that we enjoy in the United States is that it understands, and accounts for, the inherent danger of giving anyone that kind of authority over speech or press. What looks like a simple fix may very well be the lid to Pandora’s box.

• If everything’s a conspiracy, nothing’s a conspiracy by

The common thread for both Mackey and Lemon is that a single word — conspiracy — has proved elastic enough to be stretched toward a meme-maker and a journalist by administrations that agreed on little except the usefulness of the stretching.

We have to be cautious about what we call a conspiracy because expression that’s actually a conspiracy is unprotected by the First Amendment. Lawyers would describe the specific carve-out as “speech integral to criminal conduct.”

This week in

• Thank you, Brendan Carr by Bob Corn-Revere

• Violence is up and tolerance is down. What comes next? by

• Professors are being squeezed on all sides. The GUARD Act can protect them by

• Foreign journalists to face social media vetting for U.S. visas by

• When the cure is worse than the disease — FIREwire No. 41 by

• Campus hecklers are silencing pro-Israel speakers. Speech codes are not the fix by Max Friedman

Debates over Israel and antisemitism must not be controlled through censorship by administrators, governments, or mobs. Even if certain perspectives are uncomfortable or offensive to some, the First Amendment prescribes more speech as a remedy, not less. This approach allows all of us to gain a more robust understanding of contentious issues, even if we don’t ultimately change our minds. As Justice Louis Brandeis famously wrote in Whitney v. California, “Those who won our independence by revolution were not cowards. They did not fear political change. They did not exalt order at the cost of liberty.”

FIRE interview of the week

Check out FIRE Chief Counsel Bob Corn-Revere on CNN’s The Lead with Jake Tapper discussing Disney’s lawsuit claiming retaliation by the FCC against ABC.

London Calling: Ronnie’s First Amendment Roundup

• Planned Parenthood Great Northwest, Hawai’i, Alaska, Indiana, Kentucky, Inc. v. Commissioner of the Indiana State Department of Health

Seventh Circuit affirms injunction on enforcing Indiana law that bars aiding or assisting unemancipated minors in obtaining abortions without parental consent and notice insofar as it restricts accurate speech pertaining to abortion providers lawfully operating in other states.

The U.S. Court of Appeals for the Seventh Circuit affirmed a permanent injunction against the “aid-and-assist” provision of Indiana’s law governing abortion procedures for unemancipated minors that (the parties agree) make it a violation to give them accurate information about legal abortion practices and providers in other states and/or to contact, or provided referrals to, such out-of-state providers on behalf of unemancipated minors.

The court first confirmed plaintiff’s organizational standing to bring its challenge, noting that “[l]ike other corporations, Planned Parenthood has its own free speech rights” and “[w]hen … sufficiently chilled, a corporate plaintiff can establish standing to challenge a law that prohibits its speech … even when the consequences of any violation might fall most immediately and directly on the corporation’s employees rather than on the corporation itself.” The court went on to reject the state’s argument that “the private enforcement mechanism in Indiana’s aid-or-assist law means the injunction cannot actually redress or prevent the First Amendment injury,” because “Article III standing does not require a plaintiff seeking an injunction against enforcement of a statute to sue all defendants who might possibly enforce the statute against that plaintiff in the future” and a “possibility of private enforcement … does not defeat standing against members of the Indiana Medical Licensing Board.”

On the merits, the court rejected the state’s arguments that the speech at issue is integral to criminal conduct as that First Amendment exception “demands a separate underlying offense that is not the challenged restriction on speech itself,” and while “we can assume that speech aiding an unemancipated minor in obtaining an illegal abortion in Indiana would constitute speech integral to criminal conduct, … providing information and referrals to out-of-state abortion providers operating lawfully in other states does not.” “Indiana officials may disagree with the purposes behind plaintiff’s speech and with the laws and policies of other states,” the court continued, but they “cannot … make it illegal to provide information about lawful conduct in other states. Nor can they apply Indiana’s own parental consent laws to out-of-state abortions.” The court also invoked the recent Supreme Court decision in Chiles v. Salazar to rebuff the state’s argument that the aid-or-assist law cannot be enjoined because it covers more than just giving unemancipated minors information about legal alternatives and/or setting up contacts/referrals, noting that, as with Chiles’ talk-therapy-only as-applied challenge, the challenge here targeted only applications of Indiana’s law to speech that does not involve unlawful conduct.

The court next held the law is content-based applied to the speech at issue, and not viewpoint-neutral, and to invalidate that application under strict scrutiny. It began by observing the state opted to ague only that there is a rational connection between the speech restriction and its asserted interests – at too high a level of generality, without proof – noting defendants “failed to present any evidence that the state’s interests in safeguarding the parent-child relationship or protecting the physical and psychological well-being of minors would actually be furthered by prohibiting private individuals from disseminating truthful information to pregnant minors about lawful abortion practices in other states.” In doing so, the court embraced the independent right of minors to receive information, citing “a significant difference … between a parent’s right to consent to or to be notified before her child receives an abortion and a parent’s right to block access to accurate information about or a referral to an out-of-state provider of legal abortion care,” before stating: “We also cannot square defendants’ asserted interest in protecting the parent-child relationship with the strong constitutional protections given to factually accurate speech about conduct that is legal in another state.”

The court then held the law fails narrow tailoring, citing the extent to which in restricting speech it is both over- and under-inclusive. The former because the law “assumes, without showing, that Planned Parenthood intentionally provides this information to help minors evade Indiana’s parental consent requirements [whereas it] identified numerous reasons why its minor patients seek this information that are not efforts to bypass parental consent and notification.” And on the latter, “the speech at issue—information for minor patients—[is] readily available to a pregnant minor without help from plaintiff,” but rather a “minor patient herself can access it easily with a computer or smartphone” while “placing the phone call to an out-of-state abortion provider is not something only the professionals at Planned Parenthood can do.” The court also noted how “Indiana has not explained how applying this statute … where it forbids plaintiff from sharing its desired information about and referrals to out-of-state abortion providers, is sufficiently tailored to those interests where all other medical advice relating to the minor’s pregnancy is not subject to these stringent parental consent and notice requirements.”

The court accordingly affirmed the preliminary injunction, though it modified it to enjoin only Indiana’s Medical Licensing Board members from disciplining physicians based on providing information/referrals, and various county prosecutor defendants and a county juvenile court judge, but not the Commissioner of Indiana’s Department of Health (because, as plaintiffs agreed, changes in Indiana law that allowed abortions when the case started in 2017 now bar them, so the clinics are no longer licensed to perform abortions and thus cannot face licensing sanctions from the Commissioner).

Bonus case #1: In Minnesota Voters Alliance v. Ellison, the U.S. Court of Appeals for the Eighth Circuit reversed denial of a preliminary injunction on a challenge to a law that bars transmitting information within 60 days of an election if the speaker intends to impede or prevent another from exercising the right to vote and knows the information is materially false, where plaintiff seeks to opine that felons on parole cannot vote in Minnesota, even though the state’s law is “knowable and certain” that they can. The court held the prohibition is not against speech falling into any unprotected category, is obviously content-based, invalid under strict scrutiny—primarily as it “overreaches because at least one speech-promoting alternative exists: counter speech,” but also because it’s underinclusive given other speech that undermines election integrity that the state doesn’t restrict—and features a prior restraint in allowing lawsuits to “restrain a violation ... if there is a reasonable basis to believe that an individual or entity … intends to commit a prohibited act.”

Bonus case #2: In Massimino v. Benoit, the Second Circuit joined the vast majority of U.S. Courts of Appeals that have held the First Amendment clearly establishes the right to record law-enforcement activity in public (leaving only a few Circuits that have not), moving past the fact that it was not clearly established when police arrested the self-described “First Amendment auditing hobbyist” plaintiff, to deliberately articulate that it is now clearly established in the Second Circuit going forward.

International free speech stories of the week

• Azerbaijan Suing CNN for Libel: Can They Do That? (Reason) by Eugene Volokh

• Retired Police Officer Fined £1,000 Over Bacon Meme in UK (Reclaim the Net) by Cam Wakefield

The first paired “time for mass deportation” with a supposed 12-year-old migrant asking donors to move him from a three-star hotel to one with “a better halal menu, free wifi and Sky” nearer a girls’ school. The judge called that “a political issue which had been debated in Parliament” and acquitted him; the second read, “Fun facts about Bacon! People who eat bacon have a lower chance of marrying a 9-year-old!” and that one cost £1,000.

Gray said the second post was “intended to be satirical” and meant to “spark a serious debate” about Mohammed’s marriage to Aisha and child marriage in Muslim countries, The Telegraph reported. District Judge Steven Hood found it “designed to be offensive” and not capable of seriousness because it was “framed in a jokey way,” so English comedy has acquired paperwork.

• LGBTQ+ passages cut from Russian edition of Pope Francis’ memoir (TVP World) by Maria Kaminska

Videos of the Week

First up, we’ve got the latest installment in FIRE’s 1AX series: Free Speech in Donald Trump’s America | Part 1. Hear from experts like FIRE’s , NCAC’s Lee Rowland, Knight Institute’s Alex Abdo, ACLU’s Esha Bhandari, & Cato’s Thomas Berry on whether the second Trump administration has lived up to the lofty free speech promises made on the campaign trail.

Next up, enjoy the latest offering from ’s FIRE-sponsored The Blessings of Liberty podcast; a live conversation and meditation on the philosophies and aspirations of America’s Founders with law professor Robert P. George at the Chautauqua Institution.

Subscribe now

Story of the week

• Why I Quit the Tenure Track (The Atlantic) by Tyler Austin Harper

I began to develop a case of political vertigo. There I was, joining my colleagues in sifting through our curriculum like anti-racist truffle pigs; assigning an Afro Caribbean graphic novel, Marshallese ecopoetry, and the like to diversify my syllabi; and organizing a “Decolonization and Liberation”–themed MLK Day celebration meant to honor a civil-rights leader who had been assassinated while supporting sanitation workers agitating for basic labor rights. Meanwhile, the college was paying ungodly sums to corporate mercenaries to ensure that it could continue exploiting its most vulnerable employees. But if administrators felt the least bit sheepish about their hypocrisy, they didn’t show it.

This week in

• Americans left and right agree: Higher ed has an ideological-diversity problem by

• Tuskegee tells students to suit up and leave the durag at home by

• How the unraveling of Jason Arday’s career and the UK’s censorship collided by

• Pete Hegseth wants the Pentagon press corps on a leash by

• NYU’s crackdown on affinity graduations solves a problem that isn’t there by Elizabeth Maguire

• Countries are hiding censorship in ‘cybercrime’ bills. A UN treaty could make things worse by

The treaty purports to improve international cooperation in addressing crimes taking place using computer systems — but cooperation is not limited to specific online criminal acts like fraud or phishing. Instead, the treaty pushes nations to share personal data that other nations request about individuals accused of committing “serious” crimes.

But there’s an important question here. What constitutes a “serious” crime? The treaty does little to define it other than clarifying that it’s a crime carrying at least a four-year prison sentence as a punishment. For billions of people around the world, speech is a crime that warrants a long prison term.

• The romance of fair use and fiction by Ally Duong

This week in FIRE’s blog

• Student punished for ‘blasted’ comment wins court battle by

• Free speech study finds surprising similarities between ‘very liberal’ and ‘very conservative’ Americans

International free speech stories of the week

• [Canada] Justice Centre urges MPs to reject Bill C-34, warns social media law threatens privacy and free speech

Another concern raised by the organization is the creation of a new Digital Safety Commission, which would be responsible for interpreting and enforcing provisions dealing with harmful online content.

The report argues that platforms could face fines of up to $10 million or 3% of global revenue if they fail to adequately address content deemed to promote hatred, while leaving key definitions to regulators after the legislation is passed.

• France shuts mosque over Quran verses and Islamic texts used in Muslim worship

• Brazilian Influencer Convicted Of “Social Racism” After Criminally Misgendering Transgender Politician

Reminder of the Week

In light of Tyler Austin Harper’s fantastic article in The Atlantic this week (see “Story of the Week” above) and all the many problems with higher education it raised, it’s worth a reminder that and I wrote a whole damn book about the scourge of Cancel Culture inside and outside American academia.

One thing worth remembering about The Canceling of the American Mind is why I felt compelled to write it in the first place. At the time, it seemed like virtually nobody with the resources to do it was willing to undertake a serious, data-intensive examination of just how bad Cancel Culture had become, particularly in higher education, and how historically unusual the situation was.

And this was exactly the kind of work academia should have been doing itself. Universities should have been able to examine their own institutions and ask “Is there actually a problem here? How serious is it? What does the evidence show?” Instead, large parts of academia seemed barely able to keep themselves from sneering at the very idea that there was a problem—even though an enormous number of people inside higher education knew perfectly well that there was one.

The data were extraordinary. In FIRE’s 2022 faculty survey, 16% of professors reported that they had either been disciplined or threatened with discipline for their teaching, research, academic talks, or non-academic expression. 29% said they had been pressured by administrators to avoid researching controversial topics.

And the evidence has only gotten stronger. FIRE’s much larger 2024 faculty survey found that 35% of professors had recently toned down something they were writing because they worried it would cause too much controversy. The comparable number among social scientists surveyed during the McCarthy era was nine percent.

Nearly four times the McCarthy-era rate.

That is a disaster. And remember: Nobody seriously pretends McCarthyism didn’t happen.

There is an additional problem when the institutions suffering from this kind of dysfunction are the institutions responsible for producing and policing knowledge. Higher education and journalism exercise enormous influence over what gets researched, what gets covered, what gets dismissed, what gets fact-checked, and ultimately what the rest of the country is encouraged to regard as true or false. That makes failures within those institutions unusually difficult to confront, because the institutions you are criticizing have disproportionate power to decide which criticisms get taken seriously in the first place.

That was one of the reasons we wrote The Canceling of the American Mind. To my knowledge, it remains the only book on Cancel Culture that attempts anything like this combination of data, historical comparison, case analysis, and explanation of how the phenomenon actually worked.

So if Tyler Austin Harper’s piece has you wondering how American higher education got here, or if you simply want to remember what actually happened during the period when much of the press was insisting that cancel culture was imaginary, exaggerated, or merely “accountability,” check out The Canceling of the American Mind.

Subscribe now

Stories of the week

• Censorship Will Never Solve Antisemitism by

Jewish Americans have historically been among the most principled defenders of free speech. When a group of neo-Nazis planned to march through the town of Skokie, Illinois, home to many Holocaust survivors, it was a Jewish attorney who stood up for their right of assembly. David Goldberger, then the legal director of the ACLU of Illinois, was no self-hating Jew, and he certainly harbored no sympathy for the National Socialist Party of America. But he realized that “the constitutional guarantees of freedom of speech and press would be meaningless if the government could pick and choose the persons to whom they apply.”

*David Goldberger will be speaking at the gala to conclude FIRE’s Soapbox conference in Philadelphia, Nov. 4-6! Get your tickets now!

• Investigated for saying ‘word salad’!? (FIRE) by

How far can a university stretch the definition of harassment? Far enough to investigate a professor over the phrase “word salad,” apparently. Just ask Richard Himelfarb, a political science professor at Hofstra University.

• Police investigated journalist who tried to expose Cambridge professor (Telegraph) by Gordon Rayner

“I was just doing my job. Journalists should be free to ask questions of anyone, even star professors. It’s very rare that academics refuse to speak to the press – that caused me to look into this matter further and I was astonished by what I found.”

This week in ERI

This week in Expression

• Universities’ double failure on antisemitism and free speech

• How does the First Amendment govern liability for AI? by

A publisher does not become the guarantor of every idea it distributes, and a reader who acts badly on what he reads does not thereby transfer responsibility to the publisher.

*’s last day at FIRE was this past week as he starts law school in the fall, and we’re sorry to see him go but wish him the best!

• Can students learn to disagree without becoming enemies? by &

• Australia targets Telegram as Russia charges its founder with terrorism by

• College students aren’t becoming less polarized. They’re becoming less tolerant by

• Censorship never stops with just two words by

This week on So To Speak

• This week on So To Speak, FIRE EVP & host had on Michael Tigar to discuss the landmark John Peter Zenger trial and its legacy for American free speech, as well as reflections from Tigar’s six-decade legal career — from the Berkeley Free Speech Movement and the Chicago Eight to jury nullification and lawyers’ First Amendment rights.

This week in FIRE’s blog

• The real threat at Michigan’s commencement wasn’t the speech by Michelle Liao

• FIRE statement: White House’s framework puts AI evaluation in a black box

AI models are expressive tools. Their creation and use fall within the First Amendment’s protection — so the government’s efforts to regulate them rightly face serious hurdles. The public is right to be skeptical of a secret process that can be abused to reward views those in power find favorable, and punish views those in power dislike.

• University of Alabama earns FIRE’s top rating for free speech after reforming policies

London Calling: Ronnie’s First Amendment Roundup

The beginning of the end of the summer of ‘26 …

Another busy week, especially in U.S. Circuit Courts of Appeals, so we’re doing quick hits again:

• Rodríguez-Cotto v. González-Colón

Holding that the law is a content-based restriction that fails strict scrutiny, the First Circuit affirmed an injunction in the challenge by journalists to a Puerto Rico statute that criminalizes purposeful, knowing, or reckless warnings or false alarms related to an imminent catastrophe, or disseminating (including via social networks or the media) a notice or false alarm that puts at imminent risk the life, health, bodily integrity, or safety of others, or public or private property, during a governor-declared state emergency or disaster. The court rejected the government’s claim that the law is constitutional in criminalizing only knowing or reckless falsity, and thus falls in the category of historically unprotected false speech, as it is not simply a defamation law, then held that in targeting falsity and nothing more, the law is inherently content-based. And while Puerto Rico has a compelling interest in limiting falsehoods that risk public safety in an emergency, the law is not narrowly tailored as it applies not just if harm occurs but when it could, nor is it limited to speech that might be reasonably believed, and the government failed to show why the less restrictive alternative of counter-speech does not suffice. In upholding the injunction, the court ensures the concerns raised in the amicus brief FIRE joined with Electronic Frontier Foundation, Free Press, Freedom of the Press Foundation, and Pen America Center remain at bay.

• Schulte v. Leners

The Seventh Circuit affirmed denial of qualified immunity to the chair of a town board of supervisors on a First Amendment challenge to his viewpoint-based deletion of comments from the town’s website. The court held that, even prior to the Supreme Court decision in Lindke v. Freed (which the chairman’s actions predated), under clearly established law: allowing comments on a town webpage creates a public forum; websites are not unique fora to which different constitutional rules might apply; the First Amendment bars viewpoint discrimination in any forum where the government hosts speech; and the comment section could not be deemed “government speech” the town could freely censor.

• DoorDash, Inc. v. City of New York

The Second Circuit affirmed an injunction against a New York City code section that requires food delivery platforms to share with restaurants from which they retrieve and deliver food the full name, phone number, and email and delivery addresses of the customer, and the order contents, holding it violates the First Amendment’s protections against compelled speech. The court held the required disclosure did not fall within the allowance for compelled commercial disclosures under the Supreme Court’s Zauderer decision, as it was neither about the platforms’ own services but rather third parties (the customers who order the food), nor designed to counter potentially misleading commercial speech. The court then held the law is arguably content-based and subject to strict scrutiny, but even as a regulation of commercial speech it failed intermediate scrutiny. Even assuming (without deciding) economic protection of restaurants is a substantial interest that the disclosure would directly advance, the City offered no evidence the reg serves its interests better than obvious, substantially less burdensome alternatives would—especially given the reg presumes every customer consents to having their personal information shared, and requires those who want to opt out to do so on an order-by-order basis, resulting in “a marketing list that no customer asked to join and that none can be confident they have permanently left.”

• VIP Products, LLC v. Jack Daniel’s Properties, Inc.

Not a First Amendment case per se, but one in which FIRE’s Supreme Court amicus brief sought to ensure constitutionally protected parody still receives the breathing space it needs, and where the Ninth Circuit has now vacated an injunction under the federal Trademark Act’s dilution-by-tarnishment provision against VIP’s “Bad Spaniels” dog toy, which mimics Jack Daniel’s black label and bottle-shape. The court held the only Jack Daniel’s marks that are famous, as anti-dilution requires, are its name and bottle shape—and not other label elements made part of the Bad Spaniels parody, like changing “Old No. 7” to “Old No. 2”—and that the trial court erred by overlooking that parodic nature in finding a negative association between the whiskey and dog poop likely to tarnish Jack Daniel’s marks. (The opinion also admirably starts with the line “A man walks into a bar.”)

• Amazon.com Services, LLC v. Perplexity AI, Inc.

Also not a First Amendment case, but here, the Ninth Circuit vacated a preliminary injunction granted to Amazon under the Computer Fraud and Abuse Act against use of Perplexity AI products to shop for goods at Amazon.com. The court held, in part by using the Rule of Lenity, that Perplexity’s Assistant AI did not unauthorizedly “access” Amazon’s site as the CFAA requires, but rather only Amazon customers authorized by the company did so, with the AI serving as a tool for them to carry out specific acts at the website.

International free speech stories of the week

• Minns government rejects expansion of hate speech laws after review recommendations made public (ABC Australia) by Nick Dole

The Minns government has rejected key advice from a review it commissioned, which recommends hate speech laws be extended to protect more members of the community.

As a result, inciting hatred against people based on sexual orientation, disability, gender identity, religion and other characteristics will not be made illegal under the Crimes Act.

Inciting hatred based on race will remain a criminal offence.

• German high court rules comparing Israel to Nazi Germany is protected speech (Jerusalem Post)

• X says Turkish court ordered block of Istanbul Mayor Imamoglu’s candidacy account (Reuters)

Book of the Month

This month’s winner of the Prestigious Ashurbanipal book award goes to A World Appears: A Journey Into Consciousness by Michael Pollan. The book opens with my brilliant friend Christof Koch, who has spent his career chasing the neural basis of consciousness and is one of the leading champions of Integrated Information Theory. The book leaps right into the real question: what even counts as consciousness in the first place? Coherent thoughts? We’re certainly not all walking around narrating in words — not everybody has a clear inner monologue. So what’s the minimal unit? A feeling? An instinct? Something even less coherent than that, which I think we all experience anyway? And perhaps most interestingly, do plants enjoy a kind of consciousness?

I can’t do the book justice here. You have to sit with it yourself, and I think you’ll enjoy the time spent with the topic, and with Pollan.

NOTE: Christof, incidentally, co-hosts a terrific podcast with Heather Berlin called Science of Perception Box, which just opened its second season with my friend as the first guest.

Subscribe now

On July 30, the American Association of University Professors did something it has never done in its 111-year history: It endorsed a candidate for political office.

And this was not a general-election endorsement against some uniquely terrible enemy of academic freedom. No, the national AAUP waded into a contested Democratic primary for Michigan’s U.S. Senate seat and backed Abdul El-Sayed, the unmistakable candidate of the party’s left. He had the backing of Bernie Sanders and Alexandria Ocasio-Cortez, ran on Medicare for All, rejected corporate PAC money, and called for ending U.S. military aid to Israel.

He was also the more controversial choice, especially because of his position that Israel’s war in Gaza constitutes genocide and his appearances with Hasan Piker. Academic freedom, by contrast, was not a defining issue of his campaign, although El-Sayed had been an AAUP member and was strongly backed by organized labor. The AAUP proudly advertised the decision as its “First-Ever National Political Endorsement.” Todd Wolfson, the AAUP’s president, said roughly 80 percent of the Michigan members who voted supported making an endorsement and 97 percent favored El-Sayed.

Then came the less majestic denominator.

Inside Higher Ed reported that about 40 Michigan AAUP members attended the consultation call where the informal poll occurred, out of roughly 4,000 members statewide. Fewer still were on the call when the vote happened. One Michigan member said his “jaw dropped” when he saw Wolfson cite the percentages without the raw numbers.

That is not exactly a sweeping mandate, and yet here we are.

The more revealing part came when people pointed out that a century-old organization claiming to defend academic freedom for all professors might lose something by becoming an electoral participant in one party’s primary.

Wolfson’s response on X was: “Apparently our first endorsement has upset some people. Good.” He added, “AAUP isn’t going to respond by pretending politics isn’t happening. We’re going to organize.”

In another, now-deleted post, he called the critics “clowns” while straw-manning their arguments against the move.

The amazing thing is that this was offered as a rebuttal. But nobody had claimed that politics was imaginary or demanded silence about attacks on universities. FIRE has hardly been silent about the Trump administration’s abuses in that realm and others. It’s absurd to claim the issue here is some martyr-like desire to roll over, dressed as neutrality.

The deeper objection to this move, though, is actually practical. Suppose you are a professor who opposed El-Sayed, or simply does not share the AAUP’s politics, and your academic freedom is violated. Are you really going to bring your case to the organization that has declared itself a left-wing political union and endorsed the candidate you opposed? More likely, you will go to FIRE or the Academic Freedom Alliance, or look to Heterodox Academy for support.

And the damage compounds. Once the AAUP behaves like a left-wing political organization, it attracts more people who want exactly that while driving away moderates, conservatives, libertarians, and anyone else who does not want an academic-freedom organization doubling as a political faction. The membership becomes still more homogeneous in the process. Internal objections get weaker. The next partisan step seems even more obvious.

That is how an organization becomes trapped inside its own box and then mistakes the view from inside for reality.

Wolfson has been unusually explicit about what changed. At a debate in June, he said that for its first 50 years the AAUP was “a professional association representing the profession,” when neutrality made sense. Over the last 50 years, he said, it “has pivoted to being mostly a union.” Then came the key sentence: “Unions are political organizations,” and the AAUP has “shifted the organization to reflect that.”

So there it is. The AAUP’s president is saying openly that his organization has been remade into a political union. Not only that, but it has formed a political action committee and signaled that more endorsements are coming.

That may be coherent, but choices have costs.

I’ve written about the AAUP’s increasing partisanship and rejection of its founding principles before, and I didn’t do it to score points or make new enemies. I did it because this is a real loss. For most of its history, the AAUP occupied a uniquely important place in American higher education. Its 1940 Statement became part of the operating system of the modern university, and FIRE treated its best reports as models of principled analysis.

That authority depended on the belief that its standards applied across political lines. A professor did not have to vote correctly to deserve academic freedom. Once the AAUP becomes an electoral organization allied with one coalition, professors outside it have reason to wonder whether the association sees them as people to protect or opponents to defeat. And why wouldn’t they?

This past spring, I gave a presentation to FIRE’s staff about the importance of cultivating and maintaining viewpoint diversity across our ranks. Unfortunately, the AAUP has become an almost perfect case study in what happens to an organization that does not share — or worse, is outwardly hostile to — that value. What strikes me most is not simply that its leaders made a bad strategic choice. It is that they seem unable to understand why anyone could see the choice as a problem.

An organization with meaningful viewpoint diversity would have had at least one person in the room saying, “Before the national authority on academic freedom endorses a candidate in a Democratic primary, perhaps we should think about what this does to our credibility with professors who vote the other way.” And an organization with meaningful viewpoint diversity would have recognized what a seriously compromised position that endorsement puts it in.

While preparing my talk I realized there were parts of the argument I had never fully written down, and I really should, because viewpoint diversity is incredibly important. Failure to preserve it puts us where the AAUP is right now, and because of who we are and how we’re wired, we are unlikely to see it until it’s far too late.

The human problem

Human beings are self-deceiving creatures. Evolution did not design us to understand the world exactly as it is. It designed us to survive and reproduce in it. A belief that helps us do those things does not have to be true. We are perfectly capable of feeling absolute certainty about things that are, to use the technical term, bullshit.

And I do not mean “those people” are like this. I mean you. I mean me. Believing that bias is mainly a problem our opponents struggle with is, conveniently enough, one of our favorite biases.

In 1990, Ziva Kunda’s foundational paper, “The Case for Motivated Reasoning,” explained how readily the rational parts of our minds become attorneys for the conclusion we already want. Give us a reason to believe something, and we can usually construct a justification that feels principled enough to preserve both the conclusion and our self-image.

Eight years later, Raymond Nickerson’s review, “Confirmation Bias: A Ubiquitous Phenomenon in Many Guises,” showed how prior beliefs shape what evidence we search for, notice, remember, and take seriously. Evidence that supports our side gets waved through customs. Contrary evidence is detained, strip-searched, and asked to produce three forms of identification. Then, in 2001, my friend and Coddling of the American Mind co-author gave the same basic process its most memorable image in “The Emotional Dog and Its Rational Tail.” Moral intuition arrives first. Reasoning trots along behind it, explaining why our immediate reaction was also, by astonishing coincidence, the principled and intellectually serious one.

In 2017, Dan Kahan, Ellen Peters, Erica Cantrell Dawson, and Paul Slovic gave this an especially nasty twist. Participants were asked to interpret a difficult numerical table. For some, it described whether a skin-rash cream worked. For others, the same numbers described whether a ban on carrying concealed handguns reduced crime. On the politically neutral skin-cream problem, greater numeracy worked exactly as we would hope: The more mathematically capable subjects were much more likely to get the answer right. But once the identical problem was framed as gun control, political identity took over. The most numerate subjects became the most polarized. They were especially good at reading the numbers correctly when the answer supported their side, and remarkably capable of getting them wrong when the answer threatened it.

That is what makes the finding so disturbing. We like to imagine that bad reasoning is a result of low intelligence. Kahan’s study suggests that greater intelligence can instead drive us faster and more efficiently toward the conclusion we already wanted. Rather than defeat bias, intelligence actually accelerates it. In an ideologically uniform institution, the cleverest people in the room may be the most dangerous because they can build the strongest arguments for the group’s blind spots.

Then, in 2023, Aileen Oeberst and Roland Imhoff gave me the term I had been looking for: belief-consistent information processing, or BCIP. I was genuinely excited when I found it because BCIP sums up a bewildering number of cognitive biases with one simple idea — what I think of as a kind of super-bias. We begin with a belief, identity, or desired conclusion and then search for, interpret, remember, and evaluate information in ways that protect it. Confirmation bias, motivated reasoning, selective exposure, asymmetric skepticism, and a long list of other named errors are all variations on that basic move. Instead of having to memorize a zoo of biases, BCIP lets you see the animal underneath.

BCIP is the human default, which is why institutions need belief-inconsistent people.

It is not enough to have brilliant people who all think alike. A room full of smart people can become extraordinarily effective at confirming what the room already believes. That is what has happened at the AAUP. Its leaders can explain at length why neutrality is surrender, why electoral endorsements are simply another form of defending academic freedom, and why concern about political homogeneity is really just a concern that fascism is not better represented in peer review. The conclusion always seems to be that the AAUP’s current political instincts are both morally necessary and intellectually unimpeachable. What luck!

I have watched the same mechanism operate in my own work. Early in FIRE’s history, I persuaded myself that academic freedom required us to defend Elizabeth Ito, a technical-writing professor who spent ten minutes of a 50-minute class attacking the Iraq War. At the time, we had many cases involving right-leaning professors, and I wanted a visible case involving someone on the left. I also worried that turning it down would make us look hypocritical.

But the First Amendment does not give professors license to stop teaching the assigned course: In Piggee v. Carl Sandburg College, the Seventh Circuit held that a college may require classroom speech to remain germane to its educational mission, while Hardy v. Jefferson Community College and Bonnell v. Lorenzo mark the corresponding line between controversial material that genuinely serves a pedagogical purpose and speech that does not.

I was wrong, and the important thing is that my bias did not announce itself as bias. It arrived dressed as principle.

That is what viewpoint diversity pushes back against. It is not a dogmatic demand that every possible opinion receive equal weight. Many opinions are wrong. Some are embarrassing. Viewpoint diversity is institutionalized disconfirmation — a structure that helps self-deceiving people discover when they are wrong.

How consensus turns phony

The trouble gets worse in groups, and particularly in groups that think they agree.

There is nothing wrong with genuine consensus. People can disagree, test assumptions, examine alternatives, and eventually reach the same conclusion. But consensus should be an aspiration rather than an expectation. When an organization begins with the goal of consensus, it often gets groupthink instead. There is actually a fascinating timeline of research on this topic that is instructive here.

In the 1950s, Solomon Asch’s famous line-judgment experiments showed the raw power of unanimity. When everyone else in the room gave an obviously wrong answer about which line matched another, many subjects went along with it despite seeing for themselves that it was incorrect. That’s how powerful the pressure is to conform. But Asch also found that unanimity is fragile. Just the presence of one dissenter sharply reduced the pressure to conform. One person saying, “No, that is not what I see,” immediately changed the dynamic in the room.

In 1969, Serge Moscovici and Marisa Zavalloni helped establish what became known as group polarization. Like-minded people do not necessarily average one another into moderation. Instead, they often move toward a more extreme version of the position they already favored. Group members hear more arguments for their side and learn which position carries status inside the group. Cass Sunstein later pulled together decades of this work in his 2002 essay “The Law of Group Polarization.”

In 1972, Irving Janis coined the term “groupthink” in Victims of Groupthink. He argued that tightly knit groups can become so focused on maintaining agreement that careful evaluation of evidence falls by the wayside. Members censor themselves, warnings are dismissed or rationalized away, “mindguards” keep inconvenient information out of view, and silence comes to be treated as evidence that everyone is on the same page.

In 1974, Elisabeth Noelle-Neumann introduced the idea of the spiral of silence: People who think their views are unpopular often choose not to express them, reinforcing the impression that a consensus exists. Two decades later, Timur Kuran’s 1995 book Private Truths, Public Lies described a closely related phenomenon he called preference falsification, in which people publicly profess beliefs they privately reject because candor comes with consequences. The result is an institution where many individuals quietly doubt an idea while assuming almost everyone else accepts it.

And in 2004, Cass Sunstein, David Schkade, and Lisa Ellman found a comparable effect in federal appellate panels. Judges sitting with ideologically similar colleagues produced more ideologically pronounced decisions. The presence of a judge from the other political camp could change deliberation and moderate outcomes.

Numerical parity was not required. All it took was the presence of one genuinely different person.

That describes the AAUP’s evolution unusually well. First, the organization became more politically homogeneous. Then dissent became rarer. Then the absence of dissent began to look like proof that no serious dissenting position existed. Eventually, the organization moved from lacking viewpoint diversity to publishing arguments against it and publicly dismissing political disagreement as fascism failing peer review.

That is how institutions become strange without realizing they are strange.

Why actual dissent matters

Organizations often respond to the problem of bias by assigning someone to play devil’s advocate. That is a useful exercise, but it is not a substitute for having someone who actually disagrees in the room.

In 2001, Charlan Nemeth, Keith Brown, and John Rogers tested this directly in “Devil’s Advocate versus Authentic Dissent: Stimulating Quantity and Quality.” Genuine dissent produced more and better ideas than assigned dissent. In fact, in some conditions, the staged devil’s advocate actually reinforced the majority’s starting position.

This shouldn’t surprise anyone. Everyone knows the designated dissenter is doing a bit. More importantly, a role-player does not bring the same information, intuitions, priorities, or persistence as someone who sincerely believes what they’re saying. People also tend to be pretty terrible at simulating arguments they despise. Their version of steel-manning becomes, “Let us imagine that I am an ignorant bigot with contemptible motives.” Shockingly, this does not produce an especially accurate account of what the other side believes.

Dissenters have to be in the room to keep the room honest. You cannot ask a monoculture to promise that it will do better next time. The AAUP’s problem is not merely that it made a partisan decision. Its internal culture appears to have reached the point where the best response to criticism is: Yes, we are political; politics is necessary; our opponents are fascists or clowns; why would any decent person object?

That is what total lack of viewpoint diversity looks like near the end of the process.

O’Sullivan’s Curse

This brings us to John O’Sullivan’s 1989 observation that “all organizations that are not actually right-wing will over time become left-wing” — what I call O’Sullivan’s Curse.

It is not a scientific law. Institutions can plainly drift rightward. But in my experience, rightward transformation usually has an identifiable source of pressure from the top: a wealthy owner, political appointees, a government, a church hierarchy, or a board deliberately imposing a new direction.

Some of this leftward drift is organic, and some of it plainly is not. In The Canceling of the American Mind, and I spent a great deal of time on the number of different mechanisms that can produce conformity: political self-selection, professional sorting, prestige pipelines, ideological litmus tests, social-media mobs, administrative cowardice, group polarization, and the cumulative pressure to keep your head down. We called the whole obstacle course the Conformity Gauntlet.

Similarly, Neil Gross’ 2012 work emphasized occupational reputation and self-selection: Institutions known as liberal attract more liberals and repel conservatives, making the reputation increasingly accurate. His forthcoming book, What Happened to College?, is considerably less reassuring about what happens after that sorting begins. Gross describes politics seeping into every corner of campus life, bringing ideological litmus tests, self-censorship, and social exclusion. And, yes, people may simply decide they do not want a career in a place where they will spend every lunch figuratively — and perhaps literally — eating alone.

You don’t have to take Gross’ word for it either. FIRE has documented that machinery for 25 years. Over the last half-century, academics have been denied opportunities (like Mike Adams, denied promotion for his conservative writing at UNC-Wilmington), targeted by colleagues (like Dorian Abbot, the University of Chicago professor deplatformed at MIT), investigated (like Ilya Shapiro, investigated over a tweet after appointed at Georgetown), punished (like John McAdams, who Marquette attempted to fire for criticizing a colleague), or taught that professional survival requires silence (like the 55% of conservative faculty who say they at least occasionally hide their political views from colleagues). Whether coercion is the main cause or just one cause among many is almost beside the point; pretending it has no causal role is ridiculous.

The remedy has to be deliberate.

An institution does not wander back into viewpoint diversity by accident, especially after decades of sorting. It has to actually want it, value it, recruit for it, and protect the people who provide it. That means actively seeking excellent people from perspectives the organization lacks, rather than posting a job into the same professional networks and acting surprised when the applicant pool looks exactly like the existing staff. It also means bringing dissenting perspectives into conversations early enough to matter. Asking the lone conservative, libertarian, religious, or merely heterodox person to bless the finished product on Friday afternoon does not count.

Disagreement also needs rules. People should be free to attack ideas without treating every dispute as proof that a colleague is stupid, evil, or unfit for the job. Dissenters need room to speak, majorities need enough humility to listen, and everyone needs to care more about the mission than winning the argument.

The AAUP is a clean example of what happens when that fails. Its leaders do not see the drift as a problem, they see it as progress. The institution does not think it was captured, it thinks it’s seen the light. That is precisely why the remedy has to begin with a conscious decision that viewpoint diversity is not an obstacle to the mission. It is part of the mission.

The clash also has to be structured so that it produces better thinking rather than miniature cancel campaigns. People must be able to criticize political ideas sharply without treating every disagreement as proof that the other person is stupid, evil, or professionally unfit. Dissenters need enough security to speak; majorities need enough humility to listen; and everyone needs a shared mission strong enough to keep the fight focused on the question rather than the dissenter’s moral worth. That is the difference between productive conflict and Cancel Culture.

How viewpoint diversity works: disagree and commit

The thing about viewpoint diversity is that it is only effective when it’s real, and when people involved really want it. It’s a pain to have people around who are willing to tell you, to your face, that you’re wrong. It makes work harder to get done. It makes meetings more tense. But the alternative, some frictionless utopia where everyone agrees, only sounds nice. It’s actually a delusion, and delusions have consequences.

What’s required is a genuine commitment to the idea that you might be wrong, and a practice that mirrors that commitment. Before things get set in stone, ask yourself: What evidence could be presented that would change your mind? Who among you is most likely to see your blind spots? Are there dynamics at play that might skew everyone’s perception, such as the boss speaking first and then everyone else parroting her opinion?

Hiring is where this kind of commitment becomes visible. You need to bring people into the room who are able and willing to disagree.

And here I’ll quickly dispense of a ludicrous straw man of what I’m advocating for: Viewpoint diversity is not a reason to hire someone who doesn’t believe in your mission. FIRE needs people who believe in free speech as a principle. It can’t sustain itself if it hires someone who regards freedom of speech as an obstacle to the causes they care about. But mission alignment is very different from political homogeneity. We should be wary of people who cannot work well with colleagues they disagree with, who treat every institutional choice as a moralistic binary between total evil and their exact position, or who see coworkers primarily as people to convert.

That said, sometimes you have to import friction. You look for the person who makes you think, “I disagree with this person about nearly everything, but they are damned good on freedom of speech,” and then you try to get that person into the building.

FIRE’s history shows the value of those structural counterweights. I was hired as legal director in part to provide a left-leaning counterbalance to a conservative executive director. When that executive director left, I supported the conservative lawyer David French as a possible successor because I wanted to preserve ideological tension at the top. That was not just an internal voyage of personal growth — though it was. It was also organizational design.

And when an institution does not yet have authentic dissent, it should manufacture some friction while it goes looking for the real thing. Red teams, outside reviewers, adversarial presentations, devil’s advocates, and formal pre-mortems are all weaker than having an actual dissenter in the room, but they are also considerably better than nothing. Epistemic humility is a habit, and habits are borne from intentional and systematic repetition.

Another potential risk here comes from the other direction. Too much doubt and back and forth can be crippling, and organizations also have to act to further their missions. Endless deliberation stops being evidence of intellectual seriousness somewhere around hour four. That is why we, at FIRE, so often cite Amazon’s principle of “Have Backbone; Disagree and Commit.” Before the decision, argue honestly. Do not soften a real objection merely to manufacture consensus. Once the decision is made, carry it out. No pocket vetoes. No quiet sabotage. No hoping the project fails so you can perform a triumphant little dance over the wreckage. Then, let reality grade the decision.

After the fact, preserve enough institutional memory to know what the dissenters predicted. Revisit the issue if the facts change or their warnings prove correct. “Commit” cannot mean that leaders acquire infallibility the moment the meeting ends.

As I sometimes put it at FIRE: If you agree with us 50 percent of the time, you should probably be a donor. If you agree with us 80 percent of the time, you should probably work here.

The AAUP shows what happens when those habits and structures fail. The institution’s internal politics become invisible to the people inside it, because fish don’t know they’re wet. Partisanship is redescribed as principle. Old commitments are rewritten to accommodate current political objectives. Critics become fascists, clowns, or enemies of academic freedom. And an organization that once spoke with unusual moral authority becomes another predictable voice in the partisan chorus.

It can continue. It can raise money. It can organize. It can even win.

It just cannot plausibly expect anyone who isn’t in perfect ideological lock-step to trust it.

Subscribe now

SHOT FOR THE ROAD

The video below is one of those rare Shots for the Road that actually belongs on the road we are traveling. It is the staff-retreat talk from which this essay grew, and it goes further into the research and the practical work of keeping disagreement healthy.

Stories of the week

• She was fired for teaching English literature in an English class. Now she’s suing

“For 20 years, I’ve taught students how to engage with stories that excite them, challenge them, and broaden their experience of what it means to be human,” said Prabhakar. “I don’t plan on stopping now. If what happened to me becomes the norm, colleges across America will become less curious, less vibrant, and less free.”

• This European country launched a task force to target ‘hate speech.’ Now a journalist who insulted politicians is going to prison (Expression) by

There are many lessons to learn from Georgia’s continuing descent into censorship, but an especially important one is that this is simply an inherent risk of giving government officials a wider mandate to crack down on “hateful” expression: Authorities may very well take it as an opportunity to punish speech they deem hateful toward . . . authority.

This week in Expression

• How does the First Amendment apply to ‘deepfakes’? by

The current research presents little evidence that the fears about AI-generated content are materializing, or that the medium is proving more formidable at persuasion than the long line of content manipulation tools which preceded it. In fact, the lesson from those tools has been that society is fairly resilient to novel fabrications. And for the rare fabrication that is truly dangerous or damaging, the law already has existing remedies.

• The applause that disrupted nothing — yet still led to an arrest by

• The on-campus left and the off-campus right by

• The front line of combating self-censorship is the classroom by Caroline Cunningham

Peer culture is an important reason students self-censor, and classroom environments have the potential to influence this too. Conversations are carried over from the classroom to the broader campus. If students are meeting each other’s ideas in classrooms where they feel unafraid to raise their hand and express a controversial opinion, this will likely spill over into their interactions with their peers outside of the structured classroom setting. This could produce greater tolerance among students of opposing viewpoints, and could lead to a higher number of productive discussions as well as a decrease in self-censorship, as students will have learned from the example set in their classes.

• UK police review politician’s repost of guillotine art on a T-shirt by

• Why student journalists are asking to remove their names from articles by Michelle Liao

• Pennsylvania shields public universities from transparency. This student wants to change that by

• The man who wants back into San Quentin by

FIRE in the press!

• Censorship never stops with just two words (LA Times) by

Last July, the council adopted a motion to prohibit two terms — a racial epithet and a slur against women — during the legally required public comment periods of their meetings. As the rules document states, anyone who says “any variation of either of these words” during public comment will be warned. If they say a banned word again at that meeting or any subsequent one, they’ll be removed. Ejected speakers automatically are prohibited from attending future meetings for up to six business days.

This policy, championed by Council President Marqueece Harris-Dawson, is meant to push back against words he argues are “designed to attack someone’s personhood and their protected status.”

Now, you may loathe the words in question and never use them yourself. You may share the council’s desire to make public comment more civil and appropriate. You may believe Harris-Dawson and the other council members’ hearts are in the right place. But make no mistake: The council is effectively claiming the authority to ban any words it decides are inappropriate or offensive in the future. That’s a very dangerous power to grant any government official.

This week on

This week on So To Speak, host & FIRE EVP was joined by Harvard T.H. Chan School of Public Health Senior Fellow , Yale professor of radiology and biomedical imaging, , and FIRE Director of Research to discuss the Office of Management and Budget’s proposed overhaul of federal research grant rules, exploring how the changes could reshape scientific funding, academic freedom, and medical innovation.

This week in FIRE’s blog

• What student journalists need to know about social media by Dominic Coletti

• New FIRE study finds record low number of restrictive campus speech codes

London Calling: Ronnie’s First Amendment Roundup

• African People’s Education and Defense Fund v. Pinellas County

Eleventh Circuit revives First Amendment unconstitutional conditions retaliation claim based on denial of COVID-relief funds

The U.S. Court of Appeals for the Eleventh Circuit reversed the dismissal of First Amendment retaliation claims by a civil rights nonprofit serving the African community, which challenged a Florida county’s denial of the group’s access to federal COVID-relief funds for radio-station equipment to broadcast local health, educational, and emergency alerts, based on its association with the socialist Uhuru movement that promotes black community empowerment.

Viewing the essence of the First Amendment claim as being “the County manipulated the … grant program to suppress ideas that it viewed as dangerous—namely, those associated with the Uhuru Movement,” the unanimous appeals court reversed the dismissal below, holding APEDF plausibly alleged unconstitutional retaliation. While the court acknowledged “APEDF wasn’t (and isn’t) legally entitled to a COVID-relief grant,” and that “the government doesn’t engage in viewpoint discrimination simply because it selectively funds a program to encourage … activities it believes to be in the public interest, without at the same time funding an alternative program,” it is still the case, “even in the provision of subsidies,” that the government “may not aim at the suppression of dangerous ideas.”

And within that framework, the court held APEDF plausibly alleged it engaged in protected expressive association with the Uhuru movement, and that the County’s actions would likely deter a nonprofit of ordinary firmness from engaging in it. In so holding, the court disagreed that denial of what the County called “the benefit of grant funding” cannot cause objective chill simply because APEDF can “engage in protected expressive association (including with the Uhuru Movement) without the benefit of grant funding” and in fact does so. As the court put it: “The short answer is that Supreme Court precedent is to the contrary. In particular, the Court’s unconstitutional-conditions decisions … make clear that the denial of a government benefit for speech-discriminatory reasons can have a chilling effect because that type of denial can operate to penalize a speaker’s viewpoint.”

The court then held APEDF also plausibly alleged both that “the County effectively punished it for associating with the Uhuru movement—forcing it to choose between that association and an important infusion of cash that, by objective markers, it seemed to have warranted,” as well as the necessary link between the protected activity and adverse government action. The court cited text exchanges between a county board member and his aide, and transcripts of a board work session, as permitting reasonable inferences that the board member objected to funding APEDF because of its association with the Uhuru Movement, and that the board heeded that objection and revoked APEDF’s first grant and denied its application for a second because of that association.

The court thus reversed the dismissal of the First Amendment retaliation claim and remanded the case for the district court to resume proceedings. (It also reversed dismissal of equal protection claims based on the same denial of grant funding, but affirmed dismissal of a due process challenge.)

Bonus case: In Whateley v. Lackey, the U.S. Court of Appeals for the Fourth Circuit, on “a question that has divided district courts and state supreme courts in recent years: whether the character combinations on personalized license plates (or ‘vanity plates’) are government speech or private speech,” held that “Virginia’s personalized license plate messages are private speech subject to First Amendment protection,” and that the trial court erred by dismissing plaintiff’s challenge to the state’s recall of his “FTP&ATF” (“Fuck the Police & Alcohol, Tobacco, and Firearms”) vanity plates, though the court did not decide if the recall/restriction of the personalized message is unconstitutional, but rather left that to the lower court in the first instance. That outcome on First Amendment protection for vanity plate messages mirrors that which FIRE advocated in supporting U.S. Supreme Court resolution of the division in the lower courts on this issue.

International free speech stories of the week

• Australian watchdog files legal action against Telegram, saying it failed to remove violent content (AP) by Rob McGuirk & Charlotte Graham-McLay

The United Arab Emirates-based app, which reports having 1 billion users worldwide, faces a fine of up to 54.6 million Australian dollars ($38 million) if convicted in the Australian Federal Court for “alleged failure to detect and remove pro-terror material,” eSafety Commissioner Julie Inman Grant said Thursday.

The civil action is a test of Australia’s 2021 law requiring tech platforms to work to keep illegal and pro-terrorism material off their networks.

• Police arrest dozens at London rally supporting Palestine Action (Al Jazeera) by Daniel Khalili-Tari

• [Finland] Päivi Räsänen appeals conviction to European court (Helsinki Times)

• [UK] University lecturer’s dismissal appeal leaves academic freedom questions unanswered (CAF) by Frederick Attenborough

• Azerbaijan Jails Prominent TV Journalists, Media Workers (HRW) by Giorgi Gogia

An Azerbaijani court on July 27 sentenced nine journalists, media workers, and civil society activists to between 12 and 15 years in prison. It is the harshest verdict since the government began its ongoing crackdown on independent journalism.

Announcement of the Week

A quick heads-up: We’ll have more information in the coming days and weeks, but ERI plans to begin offering occasional premium pieces that will initially be available only to paid subscribers.

The overwhelming majority of ERI will remain free. But please consider becoming a paid subscriber for as little as $8 a month — especially because every dollar goes directly to FIRE!

Requiem of the Month

On this weekend of triumph, in which Spider-Man: Brand New Day became the most successful movie of the past five billion years — excuse me, 13.8 billion years — there is also a terrible crime to report: Season 2 of Wonder Man has been canceled.

As devoted ERI readers will know, I was a huge fan of Wonder Man. I thought it was wonderfully done: funny, creative, and genuinely moving in its own strange way. It reminded me of the best quirky comics I grew up with, which could come at a serious problem from such an odd or funny angle that they somehow reached something more honest.

Adding insult to injury, Wonder Man was co-created by Destin Daniel Cretton, who directed its first two episodes—and who also directed the apparently civilization-redeeming Spider-Man: Brand New Day. So, on the very weekend we celebrate his triumph, we must also mourn the destruction of one of the best things he helped create.

This is not quite a Firefly-canceled-after-half-a-season level of betrayal. At least we got a full season of Wonder Man. Still, to quote Liz Phair, it is a mark I will carry with me for a long, long time.

Read my review of Wonder Man

Subscribe now

Stories of the week

• Government withdraws subpoenas of 3 reporters for The New York Times after admitting legal errors (AP) by Michael R. Sisak, Eric Tucker, & Larry Neumeister

“Subpoenas are the last step, not the first step, but the last step,” the judge said, citing rules set to protect against violations of the First Amendment. He said the government’s actions had turned the law and the regulations “on its head.”

• Don’t let California turn universities into AI gatekeepers by &

Assembly Bill 2392 would require the California State University and the Chancellor of the California Community Colleges to convene a working group and establish procurement standards for generative AI — or GenAI — in line with certain state requirements. These requirements include ensuring that GenAI tools are evaluated for “potential harms, misuses, abuses, and bias,” and ensuring they will not produce “harmful” or “illegal” content. Although the bill provides a few examples of covered content, such as “disordered eating,” they don’t cure the underlying issue: these terms have no settled meaning and often depend on contested political and social judgments, making this mandate as broad as it is subjective.

This week in Expression

• Censoring with a vengeance: Indecent speech in the Yiddish theater by Max Friedman

Asch’s response draws on a cornerstone free speech principle: what we consider “true” or “dangerous” can change, sometimes very rapidly. John Stuart Mill wrote in On Liberty that we must embrace speakers’ right to express controversial, unpopular ideas, because those ideas may in fact be true. And even if they are not, the proper remedy is truthful speech, not censorship. Got fun nekome is a case in point. Today, the play’s rejection of racialized essentialism and embrace of same-sex relationships are points of pride for many Yiddish theater aficionados; and its radicalism in its own era inspired Paula Vogel’s award-winning play Indecent. Suppressing Got fun nekome — even assuming it might have prevented some gentiles from believing negative stereotypes about Jews — would have erased what many view as a transformative work of Jewish art.

• Shiloh Hendrix’s conviction can’t be squared with the First Amendment by

• The political bias of AI models by

• Who is trying to punish scholars for their speech? by

• Is training AI on copyrighted materials protected by the First Amendment? By

• The new war on political satire by Athena Rem

• Hawaii is a forgotten First Amendment failure by Elyse Everest

• Does MLB’s Pride Night policy violate the First Amendment?

• Flag on the play: The free speech stakes of NIL

This week in FIRE’s blog

• FIRE demands LA City Council end policy banning certain words during public comment

“A government that gets away with censorship won’t stop at just two words. It will target other speech, too — particularly criticism,” said FIRE attorney Zach Silver. “The council president himself has said he wants to ban more words in the future, and you should believe him.”

• Pennsylvania shields public universities from transparency. This student wants to change that by

International free speech stories of the week

• France becomes first EU country to ban social media access for under-15s (The Guardian)

French senators had opted for a two-tier system distinguishing between blacklisted platforms flagged as harmful to a child’s development, and those that could still be accessed with parental consent.

But the lower house’s blanket ban prevailed, despite criticism from some on the left over the age verification process, the speed of its implementation and risks of circumvention as well as privacy concerns.

• Vietnam: 5 Arrested for Ho Chi Minh Book (Human Rights Watch)

• Georgian journalist sentenced to 14 days over Facebook post (OC Media) by Mikheil Gvadzabia

• Russian activist arrested a day after predicting Putin will end up in handcuffs (South China Morning Post)

Ilya Remeslo, 42, faces charges of “spreading fakes on the Russian army”, under censorship laws Moscow introduced when it sent troops to Ukraine, according to the TASS news agency.

The charge carries a maximum penalty of up to 10 years in prison. Sergei Badamshin, named by independent media as his lawyer, said on social media that Remeslo was being transported to Moscow to face court.

[Can we add in monster tweet thread of the week? And then link to the monster tweet thread and response to Jeffrey Sack?]

TV Series of the week

I guess I technically knew that my beloved canceled show Community had a sixth season. It originally ran on NBC from 2009 to 2014, surviving cancellation scares, a famously rocky fourth season, and just about every other indignity a cult classic can endure. Then NBC finally canceled it for good… except not quite.

In one of the stranger episodes of the streaming wars, Yahoo! decided that Community would be the flagship original series for Yahoo Screen, its attempt to break into streaming. They ordered a 13-episode sixth season in 2015. The experiment didn’t exactly work out—Yahoo Screen was gone soon afterward—but somehow Community got one last chance.

I knew all of this once. Somehow I’d filed it away under “alternate timeline” and convinced myself that Season 6 either didn’t really exist or couldn’t possibly be very good.

Then Hulu added the entire series, including Season 6, and I finally gave it a shot.

The premiere didn’t do much for me. But then again, Community has always been a brilliant, gloriously uneven show. Even its pilot isn’t particularly good. Once it finds its rhythm, though, it’s unlike anything else on television.

If you’ve never watched it, it’s one of the funniest and weirdest sitcoms ever made. Dan Harmon and an absurdly talented cast somehow turned a show about a community college into a playground for formal experimentation: stop-motion animation, an 8-bit video game episode, fake documentaries, alternate timelines, action movies, westerns, and what remains the greatest paintball war in television history.

What really surprised me was that Season 6 is every bit as inventive as the best years. Rather than pretending Donald Glover’s Troy Barnes, Yvette Nicole Brown’s Shirley Bennett, or even Chevy Chase’s Pierce Hawthorne could simply be replaced, the show evolved. Keith David’s Elroy Patashnik is an absolute revelation—one of the rare late additions to a sitcom who actually makes the ensemble stronger. Paget Brewster’s Frankie Dart has become a genuine fan favorite. I liked her well enough, but she absolutely has her moments.

My favorite is the clip below, where Frankie has to deal with Dean Pelton (the incomparable Jim Rash), who has somehow become a delirious evangelist for Honda—exactly the sort of sentence that makes no sense unless you’ve watched Community.

And yes, after all these years, the prophecy is still alive. What began as an absurd throwaway line—“six seasons and a movie”—is finally becoming reality. Dan Harmon and most of the original cast are still working toward Community: The Movie. Somehow Greendale refuses to die.

Subscribe now

Story of the week

• When AI mirrors foreign censorship: FIRE statement on Oversight Board AI speech report

Authoritarian regimes are just that: authoritarian. They will lean hard on AI companies to shape and censor political speech — or perhaps even trust that those companies will voluntarily self-censor to retain market access, even without an explicit order to do so. But AI companies who profess to care about free speech have a moral duty to protect free expression and ensure their systems don’t unintentionally extend the speech restrictions of despots to users around the world.

This week in ERI

This week in Expression

• Are AI companies doing authoritarians’ dirty work? by

• The high cost of censoring speech by

• Can the government require ID before you use artificial intelligence? by

• The global crusade to ban teens from social media is now advancing across Europe by

You don’t need a weatherman to know which way the wind is blowing on teen social media bans. The bans are everywhere, and more are coming. European Commission President Ursula von der Leyen announced this week that the EU will seek some form of a youth social media ban across Europe.

• The shape of campus censorship on the left and on the right by

• After punishing people for Charlie Kirk comments, colleges are paying steep settlements by

Kirk’s assassination was a monumental moment for many Americans who considered Kirk to be an embodiment of the value of exchanging ideas with the other side. But the First Amendment does not allow public institutions to punish their employees simply because employees speak out in ways that offend others. Rather, it provides robust protections for public employees to speak as private citizens on matters of public concern, including Kirk’s death.

• America Is Failing Its Own Free Speech Standards (cross-posted from @The Next Move) by

• Michigan State’s ethics policy gags dissent — and the public will feel the effects by

• German bill would criminalize denying Israel’s right to exist by

This week on

• This week on , sat down with University of Chicago historian Ada Palmer to dig into what the history of censorship reveals about today’s free speech debates. Together they explore how censorship has evolved from the Renaissance to the age of AI, discuss the shifting relationship between state and private censorship, the role of fear and self-censorship, and why efforts to control ideas have persisted across centuries—and what that history can teach us about the challenges we face today.

FIRE in the press!

• Hochul’s data center moratorium is a disaster for New York (Poughkeepsie Journal) by

Like Europe, New York lawmakers apparently believe they can regulate energy and AI innovations in their territory without regard to the costs and consequences, and then hope they can somehow still develop their domestic technology base to meet all their future market needs. As they say, you cannot have your cake and eat it too. In reality, such moves just undermine domestic industry and options, making governments far more dependent upon surrounding jurisdictions for their energy and technology needs, while also making their citizens poorer. This is already what happened in New York following the state’s fracking ban a decade ago. Data center bans could end up being a replay of that mistake, and could also look like America’s hampering of its domestic nuclear industry over the past half century, as federal and state over-regulation made it impossible to develop critical energy infrastructure and options.

International free speech stories of the week

• Hong Kong official says booksellers should ensure titles won’t harm national security after arrests (WaPo) by Kanis Leung

But Secretary for Security Chris Tang told reporters at the legislative building that the law is clear.

“If you are a bookseller, you have the responsibility to make sure the books you sell won’t endanger national security,” he said. “It’s equal to, for example, when you are selling food, you need to ensure the food won’t cause a stomach ache and is not either poison or illegal.”

Asked if authorities would make a list of banned books, Tang said that would not be conducive to effective law enforcement targeting titles that “intend to harm the country.”

“We will not let criminals off the hook like this,” he said.

• European Commission Moves Us One Step Closer to the End of the Open Internet (Reason) by Elizabeth Nolan Brown

From posting political commentary on X to asking potentially embarrassing health questions on Reddit, reviewing books on Goodreads, posting to a forum for LGBTQ young people, consuming TikTok content, subscribing to Substack newsletters, joining a community for people dealing with psychological issues, and so much more, your real identity will be tied to your online activity. Tech companies, regulators, and law enforcement may be able to connect you to activity across the internet. There would be no more online anonymity, at least not in any real sense.

• [Germany] A Communiqué from an Author of Harmful and

Podcast of the week

On The Blessings of Liberty podcast, Senior FIRE Fellow was joined by Senior Fellow at the Brennan Center for Justice & New York Times columnist to explore the life and legacy of James Wilson, one of America’s most influential yet overlooked Founding Fathers. Drawing from his new book, The Lost Founder: James Wilson and the Forgotten Fight for a People’s Constitution, Wegman discusses Wilson’s pivotal role in shaping the Constitution, his vision of popular sovereignty and democratic self-government, and how personal scandal contributed to his historical obscurity despite his lasting impact on the American constitutional tradition.

Subscribe now

If you’re in America and smell something burning up north, it’s unfortunately not just the Canadian wildfires. In recent months, our hockey-mad friends have also been torching a number of its free speech protections — beginning with a legal shield that has been part of its law since 1970.

Until now, a Canadian charged with a hate-propaganda offence could raise the “good faith religious opinion” defense, a narrow protection for sincerely expressing an opinion on a religious subject or advancing an argument based on a religious text. The defense predates Canada’s Charter of Rights and Freedoms, which was adopted in 1982, by twelve years. The Combatting Hate Act, which received Royal Assent on June 18 and takes effect July 18, repeals it.

The government’s reassurance is that this changes nothing for the faithful: preaching, teaching, and reading scripture “in good faith” still won’t clear the bar for willful promotion of hatred.

Maybe. But two things happened at once here, and both are bad.

First, the categorical safe harbor is gone — so whether a sermon or a quoted verse is protected now depends on context, audience, and how a prosecutor reads the room. Second, in the horse-trading that got the bill through Parliament, the government also dropped the longstanding requirement that the Attorney General personally sign off before a hate-speech charge can proceed. In the absence of these two barriers, clergy and ordinary believers now get to wonder, in advance, whether saying the quiet part of their faith out loud will be read as hatred by someone with the power to charge them.

American law has a name for what happens when people must guess what they are allowed to say and, because of the uncertainty, end up not saying anything at all: a chilling effect. We have little doubt Canadians will feel that chill, even in wildfire season.

There’s more. Close behind the Combatting Hate Bill, a new Ottawa bylaw taking effect August 1 will mandate a 50-meter “bubble zone” (that’s about 164 freedom feet) around schools, hospitals, places of worship, daycares, and care homes. Inside the bubble, protest is simply banned. Not “protest that blocks the door,” or “protest that disrupts planned events,” but any protest at all. A councillor actually proposed narrowing the mandate to only prohibit conduct that actually obstructs access, and the motion lost. The Canadian Civil Liberties Association, Canada’s analogue to the ACLU, warned that two people standing on a sidewalk reading a pamphlet about peace would now be in violation of this bill. This is an intentionally broad blanket ban on expression that was once protected in Canada.

There’s more. In early July, Canadians learned of a 35-page internal memo from Industry Minister Mélanie Joly’s department — titled “Misinformation and Disinformation Strategy” — floating “legal action” against users on Facebook, X, and LinkedIn who post what the government deems “false and misleading information.” What action, exactly? Beats us. That part’s redacted. The visible portion assures everyone that whatever it is will be “documented, proportionate, and subject to senior level approval.” And don’t worry, the department itself would decide what counts as “factually incorrect, misleading or out of context.”

You read that right. This memo proposes giving the Canadian government the power to deem criticism false and punish (somehow?) the author for publishing that false information or the platform for hosting it — who exactly would be the target of “legal action” is also unclear. If that sounds familiar, it should. The Soviet Union had something very much like it. And if it strikes you as bizarre that Canada is doing this, we’re with you. Giving the government power to determine what is misinformation functionally gives it control of truth itself. Every government makes mistakes, and every government has incentives to conceal those mistakes, defend its policies, and discredit its critics. This policy and others show Canadian officials seemingly ignoring these very real risks.

One risk Canadian officials are not ignoring, though, is that of… deepfake bestiality. Yep. The recently-enacted Protecting Victims Act creates “a new hybrid offence which prohibits distributing bestiality depictions, including deepfakes, with a maximum penalty of 5 years on indictment.” We can’t help but wonder what gross (or possibly hysterical) incident led to this particular law. Canadians may want to avoid making any satirical images inspired by this Vice Presidential tweet.

As Americans, we can say with love that we are not a people known for excessive curiosity about the inner workings of our polite neighbor to the north. But after a month of speech news drifting down from our toque-wearing cousins, one of us (Adam) stood up, walked to the window, looked out over the streets of Washington, and in a moment of Seinfeldian clarity asked the open air: what is the deal with Canada?

So we tried to figure it out. We may have missed some distinctly Canadian subtlety — our legal expertise ends roughly at the border — but whatever nuance we’re missing, the overall picture is not good. Here, for similarly confused Americans, is the deal, as best we can tell.

How Canada balances expressive rights away

TL/DR: In Canada, freedom of expression is real, but conditional.

The Charter guarantees expression, religion, assembly, and association — a bundle that, on paper, maps reasonably well onto our First Amendment here in the States. The catch is the sentence right before that, which says those rights are subject to “such reasonable limits prescribed by law as can be demonstrably justified in a free and democratic society.”

Since a 1986 decision called R. v. Oakes, that clause has been operationalized as a balancing test: the government can limit a right if it has a “pressing and substantial” objective and pursues it through means that are proportionate and impair the right as little as reasonably possible.

Compared with the American system — which sorts speech by content, viewpoint, forum, and category before it ever reaches a test — Oakes looks elegant: one rule to govern them all. But that elegance comes at a sinister cost. A balancing test is also a creativity test for motivated reasoning. Give a sufficiently motivated official two values to place on a scale, and the scale will tend to tip exactly where the official already wanted it to — followed by a long, sincere-sounding explanation of why the government’s interest was simply too pressing to ignore. In other words, rather than constrain the outcome, the test can launder it.

It gets easier to censor, not harder, when the decision-maker is not a court at all. Under a pair of Canadian Supreme Court decisions known as Doré and Loyola, courts give substantial deference to professional regulators, licensing bodies, and administrative tribunals that restrict expression. Instead of reviewing the right from scratch, a court asks whether the official reasonably balanced the speaker’s Charter rights against the agency’s statutory mission. The right shows up not as a trump card but as one interest among many — and it is weighed first by the very institution that wants to limit it.

What can go wrong?

Now add the institutional setting. Many Canadian speech disputes begin not in courtrooms but in human rights tribunals — bodies designed, staffed, and funded to find and remedy discrimination. FIRE has watched this exact dynamic on American campuses with Bias Response Teams: an institution built to find harm will find harm, because that is the only tool in the drawer. It’s the hammer-and-nail problem, and it has a testable form. When an ambiguous social conflict lands in front of a body whose entire purpose is to identify discrimination, the question it asks is not “is this protected speech?” but rather “can this speech be described as discrimination?” And with the aforementioned balancing test on hand, the answer can almost always be yes.

The American equivalent, for what it’s worth, is that we have no legal doctrine at all for weighing speech against someone’s dignity — because dignity isn’t a right here, and speech is. Our resolution for the hurt-feelings case is the less formal doctrine of “Have a Coke and a smile and…”

We do not believe freedom of speech can safely be made subordinate to other rights — or to an amorphous concept like dignity — because speech is the mechanism by which a society debates, challenges, revises, narrows, and expands its understanding of every other right. And wherever governments have tried to turn “dignity” into an enforceable legal standard, its vagueness has become a very concrete threat to freedom of expression.

A cavalcade of concrete examples

So far we’ve been keeping things hypothetical for the sake of outlining principles and risks here, but we’ve got receipts too. In fact, the cases have been stacking up for years.

In 2022, Jordan Peterson’s professional regulator, the College of Psychologists of Ontario, ordered him into a mandatory “coaching” program over off-duty political posts — including calling a former Trudeau aide a name, insulting a city councillor, and mocking a magazine cover. Because Peterson is such a cultural flashpoint, many people on the left treated the punishment as a verdict on him rather than a warning about the power being claimed. But disliking the speaker is not an excuse to ignore the principle. Under Doré, an Ontario court upheld the order: a professional regulator may police your lawful political speech.

In December 2025, Professor Frances Widdowson was arrested for trespassing at the University of Victoria after attempting to speak about disputed claims of residential-school mass graves. Now, Canada’s residential-school system was unquestionably brutal, and thousands of Indigenous children died in its custody. But the famous “215 children” at Kamloops were not bodies excavated from a mass grave; they were underground anomalies identified as possible burials by ground-penetrating radar, and the claim remains unconfirmed by excavation. At another residential-school site, the excavation of 14 similar anomalies found no human remains. None of that excuses the horrors of the residential-school system, but it does show why claims should be investigated rather than treated as sacred, and why Widdowson shouldn’t be in handcuffs for discussing it.

One more: This February, the B.C. Human Rights Tribunal ordered former Chilliwack school trustee Barry Neufeld to pay $750,000 — the largest award the tribunal has ever made against an individual — for a five-year campaign of Facebook posts, emails, interviews, and public remarks against gender-identity policies in schools. Neufeld called the province’s SOGI program a “weapon of propaganda” built on an “absurd theory.” He described affirming transgender children as “child abuse,” accused educators of trying to “recruit … troubled kids into this new fad,” and warned that teaching children about sexual orientation and gender identity primed them for sexual abuse. In one particularly ugly post, he claimed there were “far more teachers who abuse kids than Catholic Priests,” and elsewhere suggested that modern schools produced more abuse victims than residential schools. So no, this was not polite policy criticism, and nobody needs to pretend otherwise. But three-quarters of a million dollars for a campaign consisting of speech, imposed by a tribunal, is a number designed to be seen from orbit. Even critics who despised what Neufeld said noticed. Pierre Poilievre called it “Orwellian.” John Cleese said he would stop booking shows in British Columbia. Neufeld is seeking judicial review.

Virtually none of these actions could survive serious First Amendment scrutiny in the United States. Our Constitution permits the government to punish threats, incitement, harassment, obstruction, fraud, and defamation — not speech merely because officials consider it hateful, offensive, undignified, or false.

That is the crucial difference: Canada asks whether the government balanced speech against its preferred interests reasonably. The First Amendment asks whether the government had any legitimate power to punish the speech in the first place. In our running effort to track speech threats worldwide, Canada keeps turning up like a bad loonie. (They stopped minting pennies more than a decade before we did.)

The convoy and the cost of emergency power

For the clearest look at what Canada will do when it decides political expression has become intolerably disruptive, consider the Freedom Convoy. And stay with us even if you thought the convoy was obnoxious, because civil liberties that apply only to protesters you like are not civil liberties. They are favors — ones that will be withdrawn if you stop agreeing with the powerful.

In early 2022, thousands of protesters and hundreds of trucks converged on Ottawa to oppose COVID mandates, especially vaccination requirements affecting cross-border truckers. The protesters believed they were resisting government coercion, defending their bodily autonomy and livelihoods, and forcing political leaders to listen to people they had dismissed. Their opponents saw an unlawful occupation that subjected residents to relentless honking, disrupted businesses and trade, and allowed a determined minority to impose enormous costs on everyone else. The protest was deeply disruptive, and related blockades closed critical border crossings, including the vital trade route at Windsor. But it was also overwhelmingly peaceful — and that distinction matters.

When organizers Tamara Lich and Chris Barber were later tried, the judge acquitted them of intimidation, emphasizing that they had repeatedly urged protesters to remain peaceful. They were convicted of mischief. The Crown nevertheless sought prison sentences of seven and eight years. The judge instead imposed 18-month conditional sentences, explaining that she wanted to deter unlawful conduct “without chilling political protest.” The prosecution’s appetite and the court’s restraint, side by side, tell you quite a lot.

But the larger civil-liberties story is what the government did in the moment. On February 14, 2022, it invoked the Emergencies Act for the first time since that law replaced the War Measures Act in 1988. Among the powers it switched on was the ability to freeze the bank accounts of people connected to the protest — without a criminal charge, a conviction, or a court order.

It’s hard to understand how extreme this measure is without putting it into the context of daily life. Imagine someone drives to Ottawa, or sends forty dollars to a crowdfunding campaign. Days later, they’re standing at a grocery checkout confused as to why their credit card charge keeps getting declined. The state has moved beyond stopping unlawful conduct and into frustrating the machinery of the ordinary life of its critics: their rent, their groceries, their savings, their ability to function. And what the state can do to one, it can do to you.

That’s not just our American independence talking. In January 2024, the Federal Court found the invocation unlawful and unconstitutional, violating the Charter’s protections for freedom of expression and against unreasonable search and seizure. In January 2026, the Federal Court of Appeal upheld that ruling. Warning against the government’s theory, the court noted that treating blockades of critical infrastructure as national-security emergencies “could stifle all kinds of protests,” including actions involving pipelines, nuclear plants, and railway lines.

This is something we keep trying to get through people’s heads as free speech advocates: The power a progressive might tolerate when it is aimed at right-wing truckers is the same power that can later be aimed at climate protesters, labor organizers, or Indigenous land defenders. Governments do not build coercive tools with a single-use setting. The tool stays in the drawer, waiting for the next official who finds your cause intolerable.

The government is now seeking leave to appeal to the Supreme Court of Canada. The Canadian Civil Liberties Association — hardly a right-wing outfit — called the invocation “unnecessary, unjustifiable and unconstitutional” and helped defeat it in court.

You’ll notice that the CCLA thread runs through this entire story. The same civil-liberties left that challenged the government over the convoy also opposed the Combatting Hate Act and Ottawa’s bubble-zone bylaw. Critics want to dismiss concern over these issues as a conservative grievance. It clearly isn’t. It is a classical-liberal warning about what happens when governments acquire broad powers to suppress expression — and the people sounding it span the political spectrum.

When consensus becomes enforced conformity

Canada is a left-leaning country, and some of its political calm comes from a degree of ideological agreement that Americans can barely imagine. Sure, there is real dissent — Quebec sovereigntism, Alberta alienation, a growing populist right — but the country’s leading political, cultural, academic, and professional institutions cluster around a consensus-minded center-left.

“Consensus” sounds healthy, and sometimes it is. A society that broadly agrees on basic rules can be stable, cooperative, and pleasant to live in. But genuine consensus has to survive disagreement. Once institutions preserve agreement by punishing dissenters, the word becomes a flattering label for conformity.

Social psychologists distinguish between “loose” cultures, which tolerate deviance and weak social norms, and “tight” cultures, which enforce norms strongly. Tight cultures often function smoothly. They are also prone to treating an unwelcome idea as antisocial, discriminatory, disloyal, or professionally disqualifying. The dissenter is no longer someone who may be wrong, but a problem the institution must manage.

That is the danger of a narrow Overton window. The people inside it may experience broad agreement and social peace. The people outside it experience investigations, compelled coaching, cancelled talks, tribunal awards, frozen accounts, and laws written broadly enough to make them guess what they may safely say. (As a Hong Kong security official explained this week, if you tell people exactly what they can’t say, that will just “let criminals off the hook.”) Canada’s consensus remains calm in part because the cost of disturbing it keeps rising.

At the height of McCarthyism, Edward R. Murrow warned, “We must not confuse dissent with disloyalty.” Canada increasingly confuses dissent with harm.

Fairness requires acknowledging that the harms prompting these laws are real. Ottawa’s bubble-zone bylaw followed genuinely ugly conduct, including a protest outside the Soloway Jewish Community Centre, which also houses a long-term-care facility. Drums and bullhorns reportedly agitated patients with dementia and retraumatized Holocaust survivors. Police-reported hate crimes in Canada more than doubled between 2018 and 2024. Lawmakers did not invent these incidents. They did, however, choose remedies that sweep far beyond them. Harassment, threats, intimidation, obstruction, and mischief were already illegal. The new powers reach the peaceful protester inside a blanket bubble zone, the preacher guessing how a prosecutor will hear a verse, and the citizen wondering whether contradicting the government’s account of events could invite legal action.

When a government responds to unlawful conduct by regulating an entire category of expression, it is targeting the debate itself. Ottawa’s bylaw even exempts labor picketing. The state, as ever, has a remarkably good ear for the difference between protest and protest it likes.

Canada is preview of the global free speech problem

Despite everything we’ve said here, though, Americans have no grounds for smugness. Many of the ideas now entering Canadian law were beta-tested on American campuses: speech as one interest to be balanced against dignity, inclusion, safety, and emotional harm; dissent recast as discrimination; censorship presented as care.

Governments across the democratic world increasingly treat expression as a social risk requiring supervision. Regulators police professional speech. Tribunals balance speech against dignity. Ministries classify misinformation. Politicians promise to protect the public from dangerous ideas. Every new authority comes wrapped in reassuring words: safety, fairness, truth, inclusion, proportionality.

Our own government has pressured platforms, pursued whistleblowers, flirted with debanking, and looked for ways to suppress speech officials considered dangerous or false. The appetite is bipartisan because power finds censorship useful regardless of who wins the election.

Canada has simply moved farther toward making that appetite routine. Its officials can invoke balance, reasonableness, professional standards, social harmony, and senior-level review. Each phrase sounds moderate in a vacuum, but together they create a system in which the government may decide what counts as hatred, misinformation, discrimination, or unacceptable disruption — and then punish people on the losing side of that definition.

The American free-speech tradition begins with distrust of anyone claiming that authority. Free speech allows us to challenge, revise, narrow, and expand every other right. It is how a pluralistic society handles deep disagreement without demanding ideological surrender. Once speech becomes subordinate to dignity, safety, inclusion, or official truth, the officials defining those values acquire power over the boundaries of public thought.

That is why Canada matters beyond Canada. It offers a preview of a democratic world with fewer outright bans and more administrative controls; fewer declarations of censorship and more balancing tests; fewer forbidden opinions and more opinions that carry professional, financial, or legal consequences.

The courts have occasionally applied the brakes, and the Canadian Civil Liberties Association keeps sounding the alarm, but Canada’s political leadership shows little appetite for changing direction. A disturbing event produces public pressure, public pressure produces broader authority over speech, and broader authority becomes the starting point for the next controversy.

Canadians have long earned a reputation for kindness, sanity, and unshowy courage. Courage is exactly what their speech culture now requires. At this rate, the anthem may need an edit: “the True North, strong and free … ish.”

A right that yields whenever officials invoke a sufficiently pressing interest is a right in name only. Canada is showing the world how censorship advances in a modern democracy: politely, procedurally, and one “reasonable limit” at a time.

Subscribe now

SHOT FOR THE ROAD

I was thrilled to sit down with @Jason Willick on ’s new “Make It Make Sense” podcast for a wide-ranging conversation about free speech in higher education. We talked about the Trump administration’s confrontation with Harvard, the use of Title IX and civil rights law on campus, and how elite universities helped create the backlash they now face. We also explored the cultural forces driving censorship, the erosion of free expression, and the high stakes of artificial intelligence, as well as how control over AI systems may become one of the defining free speech issues of the future.

Stories of the week

• VICTORY: Federal appeals court decisively rejects Florida’s ‘Stop WOKE Act’

“Today’s important decision means that college remains a place where professors and students are allowed to debate controversial topics — even if politicians disagree with them,” said FIRE senior attorney Greg H. Greubel. “Today’s ruling makes clear something we’ve known for a long time: Governments cannot censor their way to freedom.”

• The Trump administration is cracking down on dissent. That’s chilling (WaPo) by me

Streever isn’t the only American targeted by the Trump administration for criticizing its policies. The administration has aggressively pursued everyday Americans for expressing their grievances with the government. These flagrant violations of the First Amendment should frighten Americans of every political stripe, because they threaten the very heart of the American experiment: the freedom to criticize our representatives vociferously and petition our government for change.

• A Song of FIRE and ICE (The Dispatch) by Kevin D. Williamson

We’re not talking here about the lunatic ravings of some genuinely scary and dangerous figure, like maybe the author of Best Bike Rides New Jersey. This is the author of Best Bike Rides Connecticut.

This week in Expression

• Canada wants to fight misinformation, a new memo reveals. But exactly how is a mystery by

• DHS tracked a man down over an email. Now FIRE is suing by

• What’s special about poor conservative students? by

• How does the First Amendment apply to AI regulation in hiring and health care? by

“The question isn’t whether the government may police discriminatory or negligent decisions. It may. The question is what happens when the government moves beyond regulating decisions and starts regulating what’s behind them — the process and tools that feed into decision-making.”

• Democracy has a participation problem. AI may help solve it by

• Hazelwood is the training-wheels version of free speech. It’s past time the Court revisited it by

• The revolution continues by

This week in FIRE’s blog

• FIRE files lawsuit after federal agents confront New Yorker over ICE criticism

If someone is really threatening a government official, you don’t wait five months to act on it,” said FIRE Senior Attorney . “The fact that authorities didn’t respond immediately shows that David presented no threat. This pursuit is designed to intimidate lawful speech, pure and simple.”

London Calling: Ronnie’s First Amendment Roundup

• Pernell v. Florida Board of Governors of State University

Eleventh Circuit affirms preliminary injunction against Florida law barring public university and college professors from promoting or endorsing state-disfavored views on race, color, national origin, and sex

I think I can both admit my bias yet unbiasedly flag as the notable decision in our space this week FIRE’s win in the Eleventh Circuit, in which the court leans heavily into the First Amendment’s protection for academic freedom to affirm the preliminary injunction in the constitutional challenge to application of Florida’s STOP WOKE Act to the state’s public colleges and universities. The law effectively precludes instruction on specific “concepts” related to race, color, national origin, or sex, such as whether people are unconsciously biased based on race or sex, whether certain virtues like “merit, excellence, hard work, fairness, neutrality, objectivity, and racial colorblindness” are racist, and whether specific races or sexes inherently enjoy privileges or disadvantages. The federal district court in Northern Florida preliminarily enjoined the law, calling it “positively dystopian.” And the Court of Appeals has now affirmed.

The court confirms Florida’s law violates the First Amendment because “[v]iewpoint-based restrictions designed to compel or ban a set of beliefs are dangerous in any setting, and they are especially pernicious in the classroom context,” and that “goes double for broadly worded yet imprecise regulations like these, which are sure to leave both professors and their students guessing about what kind of speech might violate the rules.” Here, the Act “bars Florida’s educators from promoting or endorsing” disfavored ideas about race, color, sex, and national origin while allowing criticism of those ideas. As the majority opinion for the 2-1 panel noted: “Forcing an official government line—in a college classroom of all places—is exactly the ‘pall of orthodoxy’ that the First Amendment will not tolerate.”

In so holding, the court rejected the state’s argument that public university and college instruction is “government speech” subject to no First Amendment limits. As the court explained:

the State cast about for an existing case or doctrine that could support its speech ban in the university setting. Finding none, it tried to marry public-employee speech cases with government speech doctrine, resulting in a new rule: if the government pays a professor’s salary, it has total control over her classroom speech.

That is not a blessed union. Florida’s salary-for-speech rule is a breathtaking assertion of power to ban unpopular ideas from public discourse in the very places the State’s own statutes recognize as centers of inquiry—classrooms where students are trusted to puzzle through ideas that are good and bad, easy and hard, ideally getting ever closer to the truth. This new rule also runs headlong into the Supreme Court’s repeated, if imprecise, endorsements of academic freedom. If the First Amendment offers any boundary of protection at all for public university classrooms, this statute crosses it.

Put otherwise: “Florida’s restrictions are, as the State admits, an attempt to force uniformity of thought on students by curtailing the free exchange of ideas in universities—the very environments traditionally regarded as laboratories for expression and truth seeking.” The court made clear “the government has plenty of ways to promote its own viewpoint,” but “puppeteering every university professor in the state is not one of them.” As the court emphasized, higher education presents concepts that may challenge beliefs and sensibilities, but “the First Amendment trusts students to figure it out for themselves.”

With its ruling the Eleventh Circuit becomes the seventh federal court of appeals to expressly recognize the First Amendment rules generally applicable to public employees differ when it comes to public university faculty’s academic freedom that protects their teaching and scholarship (with no appellate court having concluded otherwise). FIRE Legal Network member Gary Edinger serving as its local counsel, and its challenge was heard alongside a related challenge filed by the ACLU, ACLU of Florida, and NAACP Legal Defense Fund.

International free speech stories of the week

• [UK] Andy Burnham’s team confirms his stance on social media ban for under 16s (Birmingham Live) by James Rodger

Technology Secretary Liz Kendall previously told the BBC that an update on VPN restrictions would be coming in July. A spokeswoman for Mr Burnham told The Mirror: “Andy is committed to keeping kids safe online, and has been vocal in his support of age restrictions for kids using social media.

“He knows this ban is a critical first step to keeping kids safe online and preventing further tragedies for families. The task now is to build on the consensus across political parties to make sure it’s enforced in full, and delivered with the urgency this issue demands.”

• [Canada] Federal memo considers legal action against social media posting false info (Toronto Sun)

• SpongeBob censored in Russia amid ‘LGBT propaganda’ fears (TVP World) by Ewan Jones

Tribute of the week

RIP to Bonnie Tyler, the singer of one of my all-time favorite songs, and certainly my favorite karaoke song. I recently had the pleasure of introducing my 10 year old to the joy of “Total Eclipse of the Heart.” 75 is far too young, but “forever’s gonna start tonight”!

Subscribe now

Stories of the week

• Frederick Douglass represents the best of us. And free speech () by

To Douglass, free speech wasn’t a side issue. It was a weapon essential to the cause. “To utter one groan, or scream, for freedom in the presence of the Southern advocate,” Douglass said in a 1854 speech, “is to bring down the frightful lash upon their quivering flesh … The right of speech is a very precious one, especially to the oppressed.” Douglass knew that free speech is essential not for those in power, but for those who are not. If the enslaved could not speak for themselves, he would speak for them. And he did so for decades as the sectional tensions continued to rise.

• VICTORY! Court dismisses Trump appointee’s speech-chilling lawsuit against Idaho conservation officer

This week in ERI

• Thomas Paine deserves more love by

By the time of his death, there was little gratitude for Paine’s contributions to the United States outside of workingmen’s associations because of his blistering attacks on revealed religion, particularly Christianity. But as America barrels toward its semiquincentennial, Thomas Paine emerges as the Founding Father Americans can celebrate without regret. Unlike his contemporaries, Paine’s radical liberalism feels strikingly modern—pro-democracy, pro-market, anti-poverty, and antislavery—and worth defending as the forces of reaction mount here at home and abroad. Without the pen of Paine, in fact, there might not be a United States to celebrate today.

This week in

• America, we’ll give our best to you by

Human beings will always be living through times that try our souls. The New World is constantly on the verge of sliding back into the dogmas of the Old. The Declaration’s signers felt the temptation to abandon principle for the sake of preservation, and we feel that same temptation today. But we cannot give in. Ever. These twin ideas — America and free speech — are so much bigger than us. Particulars of time and place, no matter how uncomfortable or disagreeable, cannot and should not destroy our faith in them. No matter how ugly it gets.

• Will Germany amend one of its most censorial laws? by

• What has FIRE been doing in the AI space? by

• 250 years later, free speech is still revolutionary by

Of course, people have been fighting for freedom since time immemorial. But what makes America unique is that this struggle was enshrined in its founding. The bedrock of our country wasn’t blood and soil, but the ideas and ideals — chief among them being the freedom to speak, to hear, and to engage in the open exchange of ideas — that make democracy possible.

This is why Philadelphia is neither a surprising nor a coincidental birthplace for these American principles. William Penn strove for religious toleration and liberty of conscience as fundamental tenets of his newfound colony of Pennsylvania. In the years leading up to the Declaration, Philadelphia became a center of colonial pamphleteering, newspapers, and political debate, renowned as the “Cradle of Liberty.”

• One Voice: Gagging education board members by

• How tolerant are disciples of Cthulhu? by

• America turns 250. Join us where it all began.

America’s 250th deserves more than fireworks. It deserves a celebration of the right to free speech that has helped make this nation great. That’s why FIRE is launching a new event right in the birthplace of American democracy itself.

In Philadelphia from November 4–6, 2026, Soapbox is bringing together headline speakers, unflinching debates, and unforgettable entertainment to explore the ideas and freedoms that will shape America for the next 250 years.

• The KIDS Act would put Washington in charge of how we can communicate online by &

This week on

• This week, FIRE EVP & Host sat down with —author of Just Plain Filthy: The Story Behind Book Banning’s Trial of the Century—to discuss the history and lasting impact of Island Trees School District v. Pico, the Supreme Court’s landmark school library book-banning case, and what it reveals about today’s debates over censorship, parental rights, and access to books.

This week in FIRE’s blog

• Early bird pricing ends July 4: Get your tickets to FIRE’s Soapbox conference in Philadelphia

London Calling: Ronnie’s First Amendment Roundup

Sestercentennial Independence Day Lightning Fireworks Round

A few quick hits and updates from a busy pre-holiday week

• Utah Political Watch, Inc. v. Musselman

The U.S. Tenth Circuit Court of Appeals reversed in significant part dismissal of a challenge to the Utah Legislature’s Capitol Media Access and Credentialing Policy—which requires reporters to be from “an established reputable news organization” and “adhere to a professional code of ethics,” while categorically excluding those from “blogs, independent … or other freelance” media—by a previously credentialed, award-winning reporter who covered the statehouse for a 25 years for institutional media, but was denied credentials on starting and becoming editor and sole reporter for his own independent newsletter. The court held the reporter plausibly alleged the policy is facially viewpoint discriminatory insofar as it allows denial of access the Legislature to reporters/media organizations for choosing to not affiliate with a larger outfit and/or to not have an editor. It also held that insofar as the credentialing policy is at least a nonpublic or limited public forum to which plaintiff had a constitutional right of access, and he averred the Legislature denied him credentials due to prior critical reporting on it, he plausibly alleged unconstitutional viewpoint discrimination. However, the court upheld dismissal of a prior restraint claim on grounds that a denial of credentials does not restrain speech but only access, and a First Amendment retaliation claim on grounds that plaintiff alleged only conclusorily that such denials objectively chill speakers of ordinary firmness. FIRE filed an amicus brief in the case, joined by Reporters Committee for Freedom of the Press.

• New York Times Co. v. Department of Defense

In an update to this case, where the D.C. federal court previously held unconstitutional and enjoined a then-new DOD policy for restrictively granted “Pentagon Facilities Alternate Credentials,” the court has now preliminarily enjoined a new policy—adopted the next business day after its earlier order—that closed effective immediately the Pentagon’s designated “Correspondents’ Corridor” from which journalists had long worked, and required going forward that all credentialed journalists are barred from entering the Pentagon altogether unless specifically invited to a press conference or prearranged interview, and even then, only if escorted by a DOD official. The court held the escort requirement violates the First Amendment because DOD issued it in retaliation for the plaintiff journalists exercising their constitutional rights, easily deeming it an adverse government action that would chill speakers of ordinary firmness, and that it was causally linked to plaintiffs’ speech given the “myriad statements by Department officials expressing disdain for reporting by The Times and other ‘legacy’ media outlets.”

• Accountability Now USA v. Greiss

An update here, too, in the case where the D.C. federal court issued a temporary restraining order against the National Parks Service/Department of Interior revoking a protest group’s permit for displaying an “8647” flag at their downtown demonstration or from otherwise seizing or ordering the flag’s removal. The court has now entered summary judgment for plaintiff and converted the TRO into a permanent injunction, holding: “Although the Court’s opinion is lengthy, that does not mean that this is a hard case. It is not. Plaintiff’s signs and flag fall well within the heartland of protected First Amendment speech, and Defendants offer no plausible basis for suppressing Plaintiff’s core, political speech.” The court’s reasoning is, indeed, straightforward:

Plaintiff’s signs accusing President Trump of raping a minor might outrage or offend some, but they do not, by any measure, cross the line from protected political speech to unprotected obscenity. That is true, moreover, even under the less demanding obscene-as-to-minors standard. Similarly, Plaintiff’s “8647” flag and similar materials, which were displayed along with other materials calling for President Trump’s removal from office, do not constitute a “true threat” or criminal “incitement” to violence. Rather, both sets of materials represent core protected speech, and, as a result, the NPS may not compel their removal or condition Plaintiff’s continuing entitlement to a demonstration permit on Plaintiff’s refraining from that speech.

• NetChoice v. Hilgers:

Another federal district court enjoins another state social media age-verification/parental-consent law, this time the U.S. District Court in Nebraska, which preliminarily enjoins the requirements that social media platforms use reasonable age verification to verify those seeking to become an account holder a on a social media platform, and to obtain “express parental consent” for minors wishing to sign up, for many of the familiar reasons, i.e., failure to satisfy strict or even intermediate scrutiny for lack of sufficient tailoring, trips the same constitutional wires as the video game restrictions for minors in EMA v. Brown, etc.

International free speech stories of the week

• Australia to double potential fines over child social media accounts

“We can all agree we would like the scheme to work better than it is currently, but that is on Big Tech taking the Mickey,” Wells told the Australian Broadcasting Corp., using an Australian slang term for deceiving, teasing or mocking.

The government announced Sunday it would introduce draft legislation into Parliament this week that would double the maximum fine to 99 million Australian dollars ($68 million) for platforms that fail to take reasons steps to prevent Australian children from holding accounts.

• Grandmother arrested under Queensland’s hate speech laws amid pro-Palestine rally outside Brisbane court

• Uganda army chief shuts down media group amid widespread curb on free speech

Book of the Semiquincentennial

’s A Glorious Liberty won my Prestigious Ashurbanipal Award back in 2021, but, frankly, it’s so good that I’m also making it my “book of the semiquincentennial.” As I’ve said before, it remains the only exception to my rule that if you’re interested in Frederick Douglass, you should read Frederick Douglass.

Subscribe now

Note from Greg: The following was originally published yesterday in Reason, but I thought it was such an eloquent tribute to one of the less celebrated, more polarizing, but incredibly influential Founders, that I wanted to make sure my readers here on ERI saw it before celebrating America’s 250th Independence Day tomorrow. Check out Reason’s July issue on “The First Free Americans,” and, if the holiday elicits a spirit of generosity or gratitude, please consider a gift to the hardest working, most reliable defender of free speech for all Americans.

On June 8, 1809, 72-year-old Thomas Paine took his last breath inside a small house in Greenwich Village. The next day the best-selling author and revolutionary’s body was loaded onto a cart and taken to his farm in New Rochelle, about 22 miles north of New York City, for burial. There was no procession, no national moment of silence, no celebration of a life fully lived.

A small group attended his funeral, including his caretaker, Marguerite Bonneville, a friend from his many years in Revolutionary France, and her son, Benjamin. As the dirt hit the mahogany coffin, Bonneville exclaimed, “Oh! Mr. Paine! My son stands here as testimony of the gratitude of America, and I, for France!”

By the time of his death, there was little gratitude for Paine’s contributions to the United States outside of workingmen’s associations because of his blistering attacks on revealed religion, particularly Christianity. But as America barrels toward its semiquincentennial, Thomas Paine emerges as the Founding Father Americans can celebrate without regret. Unlike his contemporaries, Paine’s radical liberalism feels strikingly modern—pro-democracy, pro-market, anti-poverty, and antislavery—and worth defending as the forces of reaction mount here at home and abroad. Without the pen of Paine, in fact, there might not be a United States to celebrate today.

In January 1776, Paine’s Common Sense hit the streets of Philadelphia like a cannonball. The 47-page pamphlet was an immediate sensation. Not only did Paine reject reconciliation with Great Britain and call for independence, he attacked hereditary monarchy and aristocracy as millstones around humanity’s neck. What made the text dangerous was that Paine didn’t write it for polite society. With wit and verve, he wrote it for the masses in language any farmer or artisan could understand. But Paine went further. He had the temerity to tell common people that they weren’t mules to be driven into the mud by their so-called betters. Instead, they had the right and ability to rule themselves with dignity, the divine right of kings be damned.

Paine’s democratic beliefs terrified the more elitist and conservative Founding Fathers, most notably his decades-long nemesis, John Adams. While Adams conceded that without Paine “the sword of Washington would have been wielded in vain,” he feared Paine’s egalitarian ethos would unleash anarchy into the nascent republic. Paine’s forceful argument for universal male suffrage without a property qualification petrified Adams. (Though he had a friendship with the founding feminist Mary Wollstonecraft during his time in England and revolutionary France, Paine doesn’t seem to have commented on women’s voting rights.)

A few short months after the Declaration of Independence in September 1776, Pennsylvania made good on Paine’s democratic promise. The state’s constitutional convention—presided over by Benjamin Franklin, Paine’s friend and benefactor—codified popular democracy into the state constitution while protecting civil liberties such as free speech and the right to bear arms for self-defense. In response, the embattled nation’s working people celebrated him while the colonial elite cursed him for unleashing the unforgivable conceit: equality by birth.

Paine, however, couldn’t be typecast as a typical progressive today. As the Democratic Party flirts with socialism and Donald Trump’s MAGA movement champions tariffs, Paine’s bourgeois radicalism stands firm: Markets and private property are the best ways to combat poverty. He anticipated that great wealth could be “capable of good” and rebelled against the simplistic notion that entrepreneurs and business owners were evil. “I care not how affluent some may be,” he wrote in Agrarian Justice, “provided that none be miserable in consequence of it.” In The American Crisis, Paine insisted that trade “flourishes best when it is free, and it is weak policy to attempt to fetter it,” understanding that prosperity flowed from the free trade in goods and services between individuals and nations.

Paine, however, wasn’t an apologist for the rich or inequality. Instead, he earned the admiration of working people. Not only because he was one of them, but because he fought for them wholeheartedly. Enraged by the privilege and feudalism of the Old World and any scheme to plant such poison in American soil, Paine laid out an early proposal for social insurance to provide public education for poor children, maternity benefits for new mothers, and pensions for the infirm and elderly. But Paine didn’t see his proposal as welfare. He saw it as every individual’s natural inheritance from common land being cultivated and taken out of common use. “It is not a charity but a right,” he wrote, “that I am pleading for.” Paine thought his plan would undercut the rampant inequality and dependency that corrupted the Old World.

Unlike many in the Founding generation, Paine detested slavery. When he arrived in Philadelphia in 1774, his quarters sat next to a slave market. What he called “the infernal trafic (sic)” horrified him. Later in life he would write a sentence that connected his hatred of subjugation to his celebration of democracy and individual rights: “Man has no property in man, neither has one generation a property in the generations that are to follow.”

About seven months before his death, Paine even hurled his abolitionist views in the face of former president Thomas Jefferson, according to textual analysis by the historians behind Thomas Paine: Collected Writings, published this week by Princeton University Press. In what was considered a letter of unknown authorship to Jefferson from November 1808, Paine chastised the former president and slaveholder, reminding Jefferson of the Declaration’s language he penned. “We hold these truths self-evident; that all men, (not all white men) are created equal,” he wrote. Then he came in for the kill, telling Jefferson that if slavery was so entrenched in America then that meant it was “high time for America to give up all pretentions [sic] to liberty & freedom.” Paine’s antislavery credentials shouldn’t need defending today, but they take on new salience when the U.S. government rewrites American history to sand down slavery’s wickedness.

This year, more than ever, there is an impulse to engage in hagiography when discussing the Founding generation. Paine was far from perfect, but when it comes to the questions that matter most today, he’s the Founding Father to cast our lot with. He reminds us of America’s true covenant: the right of every person to live without a master.

Subscribe now

Shot for the Road

It wouldn’t be a proper Fourth of July—let alone America’s 250th Independence Day—without revisiting the greatest patriotic anthem since Francis Scott Key. So before the fireworks (or after), treat yourself to Hulk Hogan’s immortal classic, “I Am a Real American.” You’re welcome.

Story of the week

• Tell the government: Hands off our TV stations

The American people don’t need Brendan Carr to appoint himself as the nation’s speech police. Whether the FCC is targeting The View, conservative talk radio, or your local TV station, the danger is the same: When the government exerts power over editorial decisions, broadcasters are pressured to follow only the government-approved narratives. Viewers have less choice, see fewer perspectives, and hear fewer voices.

• Cosmos Grants are back! (Cosmos Institute)

Truth is found in the open, by many people testing each other’s claims. AI can replace that contest with a single confident answer, and the disagreement that catches error begins to thin out.

• Gwar you kidding me? Secret Service reportedly investigates metal band for mock Trump execution by

Despite the anger it inspired among some Trump supporters, and despite drawing the attention of the Secret Service, Gwar’s onstage antics are First Amendment-protected speech. While true threats aren’t protected, a true threat is a serious expression of intent to commit unlawful violence against an individual. It doesn’t include hyperbole, satire, or artistic expression that simply references or depicts violence, however provocative it may be.

This week in ERI

This week in

• J. Edgar Hoover and the war on dissent by

• The ‘papers, please’ era of the internet will decimate your privacy by

• Cassius Marcellus Clay brought cannons to a free press fight by

• How does the First Amendment apply to AI? by

Far from a simple probabilistic machine, AI models are the direct result of human judgments about what information it should absorb, what values it should prioritize, what tone it should adopt, and what kinds of answers it should avoid.

• Law professors say they support free speech. Many are afraid to practice it by

This week in FIRE’s blog

• STATEMENT: ICE tracked down a New York woman who posted information she saw in a newspaper. Why her fight is everyone’s fight.

A free America doesn’t dispatch federal law enforcement agents to intimidate someone for an Instagram post of publicly available information. Free speech is the bedrock of a free society, and the First Amendment squarely prohibits ICE agents from intimidating Americans for nothing more than repeating information from a newspaper report. As we approach the 250th anniversary of our independence from England, where police now hassle residents over social media posts, let’s not follow their lead.

• VICTORY: FIRE, press freedom advocates win settlement over California law prohibiting anyone from sharing lawfully obtained information about sealed arrest records

“The First Amendment right to publish lawfully obtained information on matters of public concern is a cornerstone of an informed public, and California’s lawyers looked at the law and recognized it was indefensible under the First Amendment,” Steinbaugh said. “California’s legislature should follow their lead by rescinding this provision.”

• UW-Madison barks up the wrong tree by investigating student group for beagle rescue flier

FIRE in the press!

• Oakland County’s Adopt-a-Road controversy is a First Amendment lesson (The Detroit News) by

But if the government prohibited speech every time it made someone uncomfortable, free speech would mean little. Public colleges could shut down both pro-Israel and pro-Palestinian speakers. High schools could censor students for supporting or opposing ICE. Police could throw people in jail for posting a meme about the president. The First Amendment protects all of us from all of that. Protecting our free speech rights means sometimes standing up for speech we don’t like — not demanding it be taken down.

London Calling: Ronnie’s First Amendment Roundup

• Merriott v. City of Bossier City

Fifth Circuit revives First Amendment challenge to city council public-meeting open-comment speech restrictions all-too-frequently maintained by local bodies, such as against “personal,” “impertinent” or “slanderous” remarks

The U.S. Court of Appeals for the Fifth Circuit reversed the dismissal of First Amendment claims brought by a Bossier Parish (La.) online journalist and City Council meeting participant against a public-comment policy that purports to ban “personal,” “impertinent,” “slanderous,” or “boisterous” remarks, holding he sufficiently alleged the restrictions are unconstitutionally overbroad, vague, and viewpoint-based.

The court deemed the facial overbreadth challenge to the personal remarks ban to be plausibly alleged, holding it “captured nothing more than bold criticism of City Council members,” in contravention of the “bedrock principle underlying the First Amendment ... that the government may not prohibit the expression of an idea simply because society finds the idea itself offensive or disagreeable.” Based on “the ordinary meaning of ‘personal remarks,’” the court held, the policy “prohibits speakers from uttering an infinite number of protected, relevant statements or questions,” including: “(1) using a councilmember’s name for the record; (2) mentioning that a councilmember may have a personal stake in the outcome of a vote; (3) stating that a councilmember engaged in a corrupt act; (4) highlighting that a councilmember had recently been convicted of a crime; (5) claiming that a councilmember lied to the public; (6) suggesting that a councilmember had a conflict of interest; and (7) bringing to public attention that a councilmember had been sued—just to name a few.” In short, the court held the policy’s “possible applications are unquantifiable, especially when not harnessed by any limiting principle” and insofar as it “forbids a citizen from noting that a councilmember has—even questionably—done anything that may be relevant to the public.”

The court similarly held that in barring “impertinent” remarks – i.e., those “given to or characterized by insolent rudeness” or “not restrained within due or proper bounds especially of propriety or good taste” – the policy “runs afoul of the Constitution,” because it “is so unbounded and covers a substantial amount of core First Amendment activity in relation to its legitimate sweep.” “With no limiting principle, this provision effectively allows the City Council to decide which comments are permissible at its whim,” because: “If a councilmember is offended by comments that the Council does ‘not have the citizenry’s best interest in mind,’ or it is ‘breaking the law,’ or ‘not listening to the demands of its constituents,’ the Council could simply silence the speaker.” In the same vein, in banning “slanderous” comments, “used in the same way against the same backdrop” of New York Times v. Sullivan and its progeny, the policy “creates the same chill” on speech the Supreme Court identified, so, “standing alone, this proscription is overbroad.”

The court also holds “a standalone prohibition on ‘becom[ing] boisterous’ is too broad to be constitutionally permissible.” Under its “expansive definition” that “refers to both traditional speech and conduct, … banging one’s hand on the podium, emphatic hand gestures, pointing, crying, clapping, or simply shifting one’s tone could trigger application … as could any remark deemed ‘rowdy’ or inflammatory.” As with the personal remarks ban, the court held, “potential applications are limitless.”

The court next held the constitutional vagueness challenges to be plausibly pled, as “the terms ‘personal’ and ‘impertinent’—which are undefined—fail to provide speakers with a reasonable opportunity to know what conduct is prohibited,” with “boisterous” being vague for the same reason—violations “turn only on the listener’s discretion.” Further, the court held, “absent any explicit standards for those who apply the Policy to safeguard against its arbitrary and discriminatory enforcement by the City Council, the Policy’s other terms are likewise unconstitutionally vague.”

Finally, the court holds plaintiff also plausibly alleges the bans on personal, impertinent, and slanderous remarks are viewpoint-based, and thus unconstitutional in any public forum, including the limited public forum that city council meetings represent. Similar to the vagueness analysis, the court holds any such prohibition “turns on the perception of the individual councilmembers” and in doing so “constitutes unconstitutional viewpoint discrimination.” And because those prohibitions are viewpoint-based, “they are also content based to an impermissible degree and unreasonable” for a limited public forum. The court did, however, affirm dismissal of the viewpoint-discrimination challenge to the policy against boisterousness to the extent it can apply to speech without reference to viewpoint or content (as well as the dismissal of plaintiff’s state open meetings law claim).

The case thus returns to U.S. District Court for Western Louisiana to proceed on the plaintiff’s facial challenges.

International free speech stories of the week

• Paris Police Ban Far-Left Concert Over Anti-Police Violence Fears (NewsFire) by Dimitris Papafotis

• Japan’s Flag Desecration Bill Threatens Rights (HRW) by Teppei Kasai

• South Korea’s ‘fake news’ law tests press autonomy (DW) by Julian Ryall

Among the key changes to the law are the obligatory removal from online sites of comments deemed to be false or defamatory by the Korea Media and Communications Commission, which is also able to impose fines of up to 1 billion won (around €584,000, $690,000).

One of the most controversial additions to the law — which South Korea’s ruling Democratic Party refers to as the “anti-fake news act” — is the right to demand corrections or rebuttals to editorials or opinion pieces.

Book of the month

The winner of this month’s Prestigious Ashurbanipal Award — yes, the most prestigious of all fake literary prizes — is Nita Farahany’s The Battle for Your Brain: Defending the Right to Think Freely in the Age of Neurotech (2023). It is a terrific, urgent, and skillfully written book about one of the next great frontiers of liberty: cognitive liberty. Farahany argues that freedom of thought, mental privacy, and self-determination are not abstract philosophical luxuries. They are going to become very practical legal and cultural questions as brain-sensing technology, AI, neurotech, and behavioral prediction all get better, cheaper, and more intimate.

And while “mind reading” still sounds like science fiction, the distance between crude inference and meaningful access to our inner lives may be shorter than most people think. That’s why this book matters now. Farahany provides brilliant insight on how to translate liberty into the age of supertech — how to preserve the private space where thought, dissent, conscience, and individuality actually begin. Read The Battle for Your Brain, and pre-order her next book, Cognitive Extinction: Saving Human Thought While We Still Can. Let’s make sure we are not traipsing into dystopia while congratulating ourselves for being innovative.

Subscribe now

Regular ERI readers will know that some of my biggest hobbyhorses lately have been social media, age verification, and AI — particularly how it intersects with free expression and, perhaps most importantly, truth-seeking and knowledge creation.

Just in the last few months I’ve discussed the UK’s age-verification push, the U.S.’s own overbroad proposed AI regulations, the fight between the Pentagon and Anthropic, and much more.

In all of these posts and others, I’ve been struggling to convey how serious this is and why those of us who care about free expression are so concerned. We are in a moment where people seem ready to give as much power to the government as the government will take (which is as much as you’ll give it) to fix problems that are not nearly as bad as the proposed cure.

And to be clear, the concerns are real. People worry about what these technologies will do to work, to children, to politics, to privacy, and to society itself. Those fears are understandable. Every transformative communications technology has inspired fears, sometimes justified ones.

What worries me is that fear has a way of making old mistakes look like new solutions.

Throughout modern history, whenever a new way of communicating emerges, a familiar argument follows close behind: this technology is too dangerous, too disruptive, too influential to be left free. And the proposed answer is almost always the same — give the government more power.

But governments do not possess some magical ability to eliminate risk, panic, misinformation, extremism, or social conflict. What they can do is acquire new authority over the systems through which people communicate, learn, organize, search, publish, and exchange ideas. And once governments acquire those powers, they rarely surrender them voluntarily. And they always abuse them.

That is why the stakes are so high. We are debating the future of the infrastructure through which billions of people will speak, argue, search, publish, organize, learn, and understand the world. Resisting the ancient impulse to put authorities in charge of what people may say and know helped give us the best things in the modern world: science, liberal democracy, individual rights, pluralism, innovation, and the astonishing expansion of human knowledge. If we forget that lesson now, just as the channels of knowledge are being rebuilt all around us, the consequences could last for generations.

Today, FIRE’s executive vice president (and host of the podcast!) sent an email to staff that conveyed this sentiment better than I ever could. So I am sharing it in full here, only making minor changes and adding links and parentheticals to make it easier for people outside FIRE to understand.

From: Nico Perrino

Date: Jun 23, 2026, 1:23 PM

Subject: Current challenges for free speech & emerging tech

To: FIRE staff

Hey everyone,

At yesterday’s Rapid Response meeting1, and again today, we discussed the headwinds we face in ensuring that First Amendment standards apply to new communications technologies, particularly social media and artificial intelligence.

In Congress and the states, there is renewed momentum behind legislation that would regulate constitutionally protected speech and burden access to it through age-verification and other requirements.

In the courts, recent rulings from the Fifth and Sixth Circuits involving laws passed in Texas and Ohio call into question decades of established First Amendment precedent.

And in the court of public opinion, social media and AI remain deeply unpopular. Many people do not think of free speech when they think of these technologies, and they may not instinctively understand how our arguments apply.

We are already doing a lot to meet these challenges: directly litigating or filing amicus briefs in nearly every major case, engaging with lawmakers, hosting a free speech and emerging technologies event series, coordinating through an internal working group, communicating with our partners, and staffing up to accommodate the expanding workload.

That said, the headwinds are growing stronger. At yesterday’s Executive Team meeting, we agreed that an important first step is to acknowledge the scope and scale of the opposition. From there, we need to ask what more we can do, and how we might adjust our tactics and strategies to be more effective.

Over the coming months, you can expect more conversations across departments toward that end. We want each department to think critically and creatively about what it can do to help ensure that First Amendment protections extend to new communications technologies. We also want to improve our messaging by stress-testing our talking points internally and externally, and sharing which arguments are landing and which are not.

I recognize that these questions are hard. Many concerns about social media, AI, and other emerging technologies are real and deserve thoughtful engagement. Civil liberties work often requires us to acknowledge that the case for censorship can be emotionally and politically powerful.

Sometimes the harms cited are grave: war, national security threats, civil unrest, child safety, or life-and-death emergencies. But the old solution is still the wrong one: giving the government more power over speech, thought, and knowledge itself.

Our role as a First Amendment organization is to insist that, even when a technology is unpopular or disruptive, the government must still meet the demanding standards the First Amendment requires before it regulates speech.

Ultimately, we want to look back at this moment 25 years from now and say, “We did everything in our power to meet the moment.”

Civil liberties victories can take a long time, and the opposition can be intense. It took 140 years for the Supreme Court to first strike down a speech restriction on First Amendment grounds. Comstockism, McCarthyism, and censorship during World War I were all astoundingly popular.

The same pattern has often held for new communications technologies. We dealt with centuries of licensing requirements for printed matter. The Supreme Court initially held that movies did not receive First Amendment protection. Broadcast communications have been burdened by censorial laws for nearly a century.

The internet was the major exception. When it began to be widely adopted in the 1990s, the government quickly tried to regulate its content by likening it to broadcast communications. Public opinion was largely on the censors’ side, and the Communications Decency Act passed with overwhelming majorities in both houses of Congress.

Fortunately, civil libertarians recognized what was happening and acted quickly. The Electronic Frontier Foundation had formed. The Cypherpunks got organized. And the ACLU developed a litigation strategy that culminated in Reno v. ACLU, which applied full First Amendment protection to the internet. EFF attorney Mike Godwin thought he might be long retired before the Supreme Court vindicated First Amendment rights online. But when the Court delivered its ruling in 1997, he then asked himself “whether I ought to retire from civil liberties work, my job being mostly done.”

Of course, his job would never be done. That’s not how civil liberties work. As ACLU founder Roger Baldwin often reminded people, defending the Bill of Rights requires “eternal vigilance.” That’s why Greg calls free speech “the eternally radical idea.” There will always be censors, and there will always be people — sometimes most people — who believe they have a point. Oftentimes they do. But after the passions and factions of any given moment subside, we almost always look back at censorship as a misguided solution, no matter how commonsensical it seemed at the time.

When I interviewed former ACLU Executive Director Ira Glasser for my forthcoming book, I asked him why the ACLU fought for free speech rights in the early internet era. He told me, “If the First Amendment wasn’t going to apply to this new means of speaking in roughly the same way that it applied to the old means of speaking, then free speech was going to really be wounded, because in a short period of time that was going to be the major way in which speech was spoken. And that turned out to be an understatement.”

I think we stand at a similar moment today. Whether we like it or not, much of where we speak and access speech will occur through new internet-enabled technologies, including social media and AI. For many people, these technologies are already a primary way of gathering news and information. If the First Amendment does not apply to them in the same way it applies to older technologies, then the First Amendment — and FIRE’s mission to defend free speech — will be seriously compromised.

We cannot let that happen.

If you’ve made it this far, thank you. For now, please begin thinking about creative tactics and strategies your department could deploy on these issues, and stay tuned for more scheduled discussion in forthcoming meetings.

Onward —

Subscribe now

SHOT FOR THE ROAD

In light of Nico’s points — particularly regarding the fact that the technologies are new but the proposed solutions are the same — I’d like to reshare my remarks at FIRE’s 2023 gala in New York City, which I’ve titled “Fight the Guardians.” Power will always tell us it needs still more power for our own good, and it’s up to those of us who believe in the truly radical idea of human freedom to push back.

Story of the week

• Statement on shutdown of Anthropic’s Fable and Mythos

By demanding Anthropic eliminate access solely for foreign users — a functional impossibility, per the company — the government is exercising the kind of “kill switch” on artificial intelligence we have seen repressive regimes abroad use on the internet.

By leaving the public in the dark as to the specific threat identified and the statutory authority invoked to address it, the government’s dramatic action functions as an arbitrary abuse of power. The export controls serve as a prior restraint or a licensing system, requiring Anthropic to receive government approval as a condition of operating expressive systems.

This week in ERI

• The online ‘safety’ trap by me

Come for the AI-free speech analysis, stay for the Woodrow Wilson takedown!

This week in

• Why do so many law professors believe one thing, but teach another? by

• Ida B. Wells: Journalist, activist, civil rights icon, and free speech hero by

• What does AI have to do with the First Amendment? by

Applying First Amendment principles to AI-generated expression is both a natural extension of those principles and a necessary step to prevent those principles’ weakening in other applications, including in areas like newspapers and video games.

This week on

• This week, Nico spoke with activist and self-described Satanologist about using satire, litigation, and “malicious compliance” to test the limits of religious liberty, free speech, and government viewpoint neutrality under the First Amendment.

This week in FIRE’s blog

• Outside a New Jersey immigration detention center, reporting on protests may cost you a hospital visit by Athena Rem

• Victory: Wisconsin school board backs away from ban on criticism during public comment by Isabelle Brito

On May 21, FIRE wrote to the board, explaining that the restrictions were unconstitutional and urging the board to rescind them or confirm they no longer reflected board policy.

The board responded, confirming the statement “does not reflect the Board’s current formal or informal policies and practices” and that it has “no plans to . . . enforce such restrictions in the future.”

That’s a win for every resident who wants to hold local officials accountable. At government meetings, the First Amendment protects more than praise and flattery.

• UK teen social media ban is anonymity-killer for adults by

• New York wants warning labels on AI in the news by

FIRE in the press!

• Brits May Soon Say Goodbye to an Anonymous Internet (Persuasion) by

But there’s a problem: these “under-16” policies will blunt all UK citizens’ ability to speak freely and anonymously online. All users, whether they’re 15 years old or 55, will need to prove their age in order to post on social media. This raises significant privacy concerns, not just for users’ ages, but for their identities. The process of age verification will cut away vital firewalls around users’ anonymity. As the rush to ban children from social media hits a global fever pitch, it’s well past time we abandon the faulty claim that only children will feel its effects.

London Calling: Ronnie’s First Amendment Roundup

• Borecky v. Nassau County

NY federal court preliminarily enjoins ban on expressive or symbolic conduct within 35 feet of house-of-worship entrances or driveways, or within 10 feet of anyone in a 100-foot bubble around the entrances without their consent, for one hour before through one hour after any service, meeting, or ceremony

Calling it “unprecedented,” the U.S. District Court for Eastern New York preliminarily enjoined Nassau County’s Religious Safety Act banning picketing, advocacy, literature distribution, and all other expressive or symbolic conduct within 35 feet of entrances or driveways of houses of worship, along with its related ban, applicable in a 100-foot bubble around the entrances, against getting closer than 10 feet to anyone for such expressive purposes without their consent. Under the RSA, houses of worship are “any church, synagogue, mosque, temple, or other … location regularly used primarily for religious worship, religious education, or religious services,” an “entrance area” is “any doorway, threshold, entryway, gate, ramp, or other point of ingress or egress,” and a “driveway” is any “entry from a public street to any parking lot” where the house of worship has “an ownership, easement, leasehold interest, or other property right.” The court agreed with the plaintiffs challenging the combined bans, which apply an hour before though one hour after any religious service, community meeting, ceremony, or congregational, educational or organizational meeting or event, that: “the law’s benign purpose—to allow individuals to freely attend religious services and other events at these places, without harassment or fear, and to ensure public safety—is inconsistent with the broad sweep of the law and constitutes a substantial infringement on core First Amendment activity.”

That was so given the RSA’s practical effect. The court cited evidence that there are nearly 1000 places of worship in the county. So, with the 35-foot buffer and 100-foot bubble resulting in restrictions that reach public streets and sidewalks, the combined effects of density of places of worship and RSA limits make entire public streets off limits for any of the prohibited conduct—which the court notes “cover the waterfront” of the First Amendment activity. And given the plethora of religious services, community meetings, and other events at houses of worship, and that the RSA covers non-religious activities—i.e., any community meeting or event—its bans can make prohibited zones off-limits, based just on weekly scheduled events, for upwards or 12 hours a day, and more than 16 hours a day in some places.

Noting the parties agree the RSA is content-neutral and thus subject to intermediate scrutiny, and that insofar as its operation reaches traditional public forums it must, in addition to advancing an important government interest and not burdening substantially more speech than necessary, leave open ample alternative channels for communication, the court held the 35-foot buffer likely cannot withstand First Amendment scrutiny. It held (and plaintiffs did not contest) that protecting public safety and the right to religious freedom are important government interests. But the buffer likely flunks narrow tailoring given it “imposes restrictions on such speech essentially without limitation as to time, and in some places over sidewalks and public streets.” So the “combination of these effects imposes nothing less than severe burdens on Plaintiffs’ activities,” while “Nassau County’s interests in protecting religious freedom and public safety are insufficiently linked to the imposition of such broad suppression.”

“If the goal is to avoid harassment, intimidation, violence, or threatening speech,” the court continued, “the County could have drafted a law that criminalized such conduct. It need not have also banned peaceful conversation, polite exchange, and information distribution on public streets—what amounts to the extreme step of closing a substantial portion of a traditional public forum to all speakers.” And the county’s “proposed alternatives—that Plaintiffs demonstrate 35 feet away, leaflet during the hours there are no services, or stand stationary within the 100-foot bubble zone—are inadequate.”

Similarly, the bubble governing the 10 feet surrounding each person within 100 feet of house-of-worship entrances was not narrowly tailored, especially given that, combined with the 35-foot buffer, it is quite different from Hill v. Colorado’s 8-foot buffer around individuals within 100 feet of healthcare facilities. And the county “failed to articulate any rationale for having both a buffer and bubble … to advance their interests in protecting religious liberty and public safety.”

The court also found “an independent basis that makes the RSA susceptible to invalidation,” in “the unbridled discretion of an officer to police violations of the law.” Speakers face criminal liability under the RSA after they “receive a verbal, written, or other communicative or expressive instruction, warning, or order” from an officer—but “given the breadth of the expressive conduct and speech implicated by the law, an officer is left with the discretion to determine whether the individual standing in silence wearing a t-shirt with a political message is violating the statute or not engaging in expressive conduct at all.” Relatedly, the law is standardless as to what “instruction, warning, or order” will suffice, because the law “could encapsulate anything from a loud and clear instruction to disperse or risk criminal liability to a simple raising of the arm and pointing down the street.” In short: “There are no rules, standards, or guidelines to determine what constitutes conduct that violates the law.”

Having held plaintiffs likely to succeed on the merits of establishing the RSA is unconstitutional, the court held they also established irreparable harm, and that the public interest favors enjoining the RSA, noting that, even if the county “has a powerful, ongoing interest in protecting its residents from harm when attending religious service,” “an important government interest that is insufficient to demonstrate constitutional validity, cannot tip the equities or public interest in the County’s favor.” The court thus preliminarily enjoined the bans in their entirety and their enforcement against anyone.

International free speech stories of the week

• Telegram challenges India ban over exam paper leak fears (BBC) by Nikita Yadav

• Canada tables bill restricting social media, AI chatbots for under-16

“The evidence is clear: online harms are putting our children especially at risk,” said Anna Gainey, Secretary of State (Children and Youth). “The Safe Social Media Act will hold platforms accountable and help make the internet safer.”

Specifically, the framework will operate through three core duties: the duty to protect children, which will apply to all regulated services under the act; the duty to act responsibly, which will require social media services to mitigate risks associated with exposure to harmful content, apply labels to synthetically generated content, and provide clear and accessible ways for users to flag harmful content and block other users; and the duty to make certain content inaccessible, which will require the rapid removal of content that sexually victimizes a child, or intimate content communicated without consent, including deepfake sexual images.

• (India) ‘Stand-up should be banned’ says city mayor (Chortle)

Video of the week

The great Bob Odenkirk reprised his iconic role as the hilarious, if perhaps a bit unscrupulous, attorney, Saul Goodman, from Breaking Bad and Better Call Saul to wish America a happy 250th birthday. It made its rounds on social media, but in case you missed it!

Subscribe now

On June 12, the federal government ordered the AI company Anthropic to cut off foreign-national access to its newest and most powerful AI models, Fable 5 and Mythos 5. Anthropic said the directive, issued under national security authorities, forced it to disable those models for everyone. According to The Washington Post, officials initially gave Anthropic 90 minutes to take one of the models offline.

Maybe the government had good reasons. The public does not yet know enough to say, because the directive is still not public. And that is itself a serious problem. Without identifying the specific threat being addressed or the statutory authority being invoked, these controls look less like a narrowly tailored national-security measure and more like an arbitrary infringement on the expressive rights of Anthropic and its users. AI models are expressive tools: They reflect choices about training, weighting, design, and guardrails. In users’ hands, they enable people to gather, create, and share information. The danger is not only that a particular model gets shut down. It is that the government claims a kill switch over AI systems — the power to suppress models that contradict its preferred views, embarrass its officials, or come from developers who defy or oppose the administration.

But while the Anthropic fight has been taking up all the oxygen and attention, two other online-safety stories have been unfolding that deserve at least as much attention.

In Britain, Prime Minister Keir Starmer announced that the government plans to ban children under 16 from major social media platforms by spring 2027. The proposal would require big tech companies to introduce age-verification technology, building on the UK’s Online Safety Act and following the model of Australia’s under-16 social-media ban. The UK government says the goal is to protect children from online harm. But critics immediately pointed out the obvious problem: To keep 15-year-olds out, platforms must find a way to check who is 15. And to do that at scale, the internet starts moving toward ID checks, face scans, and third-party verification systems, or some functional equivalent.

Meanwhile, at the White House, offices connected to the chief of staff, the first lady, the Office of Science and Technology Policy, and the National Economic Council reportedly met with children’s online-safety groups. The grand bargain forming in Washington is that Congress may block states from adopting or enforcing some of their own AI laws — what’s known as “preemption” — but only as part of a broader package of federal internet regulations. Those rules could pressure platforms to remove more speech, require age checks that double as identity checks, create new takedown rights over AI-generated images or voices, and assert government control over how AI systems are designed and what they may say.

And like the Anthropic order, it’s all in the name of safety.

It’s important to point out here that I am not a social media utopian. As regular ERI readers and FIRE supporters will know, I co-authored The Coddling of the American Mind with social psychologist . In that book, we argued that overprotection was making young people more anxious, fragile, and less prepared for adult life. I also happen to think that the phone-based childhood has been a bad bargain for many kids, and that phones should be out of schools entirely.

But that does not mean governments should build a national identity-checking system for the internet, give government a foot in the door for controlling artificial intelligence, or create broad new liabilities that pressure platforms to suppress lawful expression in the name of protecting minors. I have disagreed with Jon on this point, as I’ve outlined in my review of his most recent book The Anxious Generation. And I certainly disagree with his support for the UK’s under-16 social-media ban. Britain is one of the last countries on earth that needs another lever for controlling speech — or another excuse to make anonymity online functionally impossible.

The Anthropic order, the Washington bargain, and the UK age-verification push are powered by the same dangerous assumption: if a problem can be described as a matter of safety, the government should be trusted with new leverage over speech.

The slope is more than slippery. It’s a mudslide

Unfortunately, something I and other free speech defenders have to repeat until we’re blue in the face is that regulation rarely ends where it begins. Lawmakers start with a sympathetic target, a modest-sounding standard, and a category of speech almost nobody wants to defend. Then the category expands, the logic of control migrates, and before long, a tool sold as a scalpel becomes a warhammer.

One big problem with the discourse surrounding all of this is that age verification is something of a misnomer. This is because the systems necessary to make that kind of screening possible also inevitably make anonymity for adults impossible to preserve. That means building a “show me your papers, please” internet at the very moment when anonymity is most essential.

What people arguing in favor of these policies don’t seem to get is that anonymity is not a loophole in free speech. In fact, it’s one of its oldest protections. Anonymous and pseudonymous speakers have included dissidents, whistleblowers, abuse survivors, religious minorities, political minorities, and citizens with something unpopular to say. The Federalist Papers were published under a pseudonym. So were anti-slavery pamphlets, samizdat writings under Soviet rule, and the anonymous tips and leaks that exposed corruption, abuse, and illegality in institutions that wanted those facts kept hidden.

Now is a reckless time to weaken that tradition. Freedom of speech is in global freefall. Germany has used hate speech, insult, and anti-propaganda laws to punish satire, political memes, and discussion of crime statistics. Britain reportedly arrests thousands of people a year — 13,000 in 2024, according to The Telegraph — under online-communications laws, including for social-media posts and private messages deemed grossly offensive or menacing. The European Union pressures platforms through sprawling digital regulation. In much of the democratic world, “safety” has become the respectable language of speech control.

The United States has long been the great exception, but that exception looks shakier by the day. On campus, administrators have too often tolerated or facilitated shoutdowns, heckler’s vetoes, bias-response systems, and investigations into protected speech. In Trump’s Washington, threats against universities, law firms, media companies, political opponents, and even private associations have become routine tools of governance. Groups on the left helped normalize informal censorship as moral imperative; many on the MAGA right are learning to enjoy formal censorship as retribution.

Another issue is the overextension of certain terms to meet ideological ends. The most dangerous word in this entire debate may be “harm.” On campus, we at FIRE have spent years watching this term expand from actual threats and harassment into reading controversial books, inviting visiting speakers, making jokes, offering classroom hypotheticals, hearing different political opinions, and even entertaining the presence of disfavored arguments. That was one of the central warnings of Coddling: When institutions teach people to treat emotional discomfort as danger, censorship starts to look like compassion.

Last year, then-attorney General Pam Bondi echoed this logic from the right, treating critical comments about Charlie Kirk after his murder as grounds for official retaliation. Practically anything important enough to argue about can be repackaged as harmful to someone, and “harmful to children” is the most powerful and manipulative version of the claim. That is a perpetual source of frustration for those of us who actually do want to protect kids: child protection can become the skeleton key that opens the door to censoring adults.

Some threats sound scary and new, but they’re old — and already protected against

The NO FAKES Act, designed to allow people to control the use of AI-generated versions of their voice or likeness, is the easiest part of this package to sympathize with. Nobody wants their voice or face cloned for a scam, fake endorsement, or humiliating sexual image. But existing law already protects us against fraud, defamation, privacy law, right of publicity, extortion, or harassment. AI will create hard cases, for sure, but courts can develop narrower rules around actual abuse in light of this new technology. A broad new federal right over “digital replicas,” especially if enforced through notice-and-takedown measures, will inevitably sweep in things like parody, satire, political commentary, documentary work, memes, reenactments, and criticism. Another issue is the vague and overbroad wording of the proposed law. The exceptions offered in the bill are so narrow that speakers cannot reliably determine in advance whether they apply. When in doubt, it’s smart to keep quiet, which is exactly how chilling effects work.

I will say here that federal preemption of state AI laws is, by itself, a very good idea. States are already producing a thicket of inconsistent, vague, and sometimes flatly asinine AI rules that aren’t compatible with the breadth of expressive rights Americans enjoy.

But the answer to state-level overreach is not federal overreach. As we speak, AI is becoming a tutor, editor, research assistant, librarian, debate partner, and search engine — something close to the future operating system of the planet. Rules that govern what AI systems may say, how they must be designed, what risks they must avoid, and what answers they must suppress are automatically and unavoidably rules about the future of knowledge creation itself. This is why FIRE and the have teamed up to create and promote truth-seeking AI.

Our goal must be to keep control of AI out of the hands of the powerful

The great liberal achievement of human history was separating truth from power: Denying kings, priests, commissars, bureaucrats, and mobs the authority to decide what everyone else is allowed to know. Power always wants that authority back. It wants to decide which claims are false, which harms matter, which depictions are unauthorized, which jokes are too risky, which AI answers are unsafe, and which anonymous speakers must be unmasked.

We should do everything we can to keep power’s hands off the systems through which humanity will increasingly think, learn, argue, and discover.

Granted, there are real concerns about kids and the effects of technology on their development. But a hard social problem does not justify building permanent speech-control infrastructure. A phone-based childhood can be improved through cultural adaptation, parental supervision, better parental tools, and getting phones out of schools. A federal system that weakens anonymity, pressures platforms to remove lawful speech, and gives government leverage over artificial intelligence is unlikely to solve the problems it ostensibly wants to solve, and introduces a different order of danger in the process.

The FCC is an example of a solution that has turned into a problem. It was created in large part to solve a practical issue of its time: allocating the electromagnetic spectrum. However, again and again, it has threatened First Amendment values — and under current Chairman Brendan Carr it has shown how quickly a communications regulator can become a tool for jawboning (using government pressure on private actors to do what the government cannot directly command) media companies, threatening licenses, and punishing perceived enemies of the president.

That is why civil libertarians sound so unreasonable and callous at the outset of these kinds of arguments. It’s not that we don’t see the concerns animating people. It’s not that we don’t care about addressing those concerns — in fact, most of us care a lot. What we’re doing is looking beyond the stated purpose of the law and identifying the agency that will inevitably be built, the precedents that will undoubtedly be set, and the politicians who will enthusiastically inherit and wield the tool to serve their own ends.

As Justice Robert Jackson wrote in West Virginia State Board of Education v. Barnette, the First Amendment was designed “to avoid these ends by avoiding these beginnings.”

That’s what I’m urging all of you — especially those of you with children, who are concerned for their safety and their future: Think like an American, in the best tradition of our 250-year argument over liberty. Imagine every new power in the hands of your worst political enemy, because eventually, it will be. Try to consider how those enemies might use and abuse the power you’re so eager to grant your allies. And when you come to the conclusion that this is a terrifying state of affairs, come back to this moment and recognize that you can stop it from ever becoming reality.

Subscribe now

SHOT FOR THE ROAD

I knew Woodrow Wilson was terrible before it was cool.

I remember studying Wilson in high school and thinking, even then: Wow, this guy had a ton of terrible ideas.

It wasn’t just that he kept the United States out of World War I too long — a decision that may have made the war longer, bloodier, and more destabilizing than it needed to be. It was also that so much of his worldview seemed to mistake abstractions for reality. Did he really think the cause of World War I was some seminar-room equation about diplomacy and ideals? Germany wanted its war with France before Russia became too powerful. Austria-Hungary wanted Serbia punished and subordinated. Nationalism, empire, militarism, alliance politics, and old-fashioned power-seeking were doing what they usually do. And Wilson’s grand solution — national self-determination — sounded noble until you noticed that, if followed consistently, it could leave humanity trapped in an endless war over borders, peoples, language, ethnicity, religion, history, and grievance.

And, of course, didn’t people know he was a total racist?

On that last point, at least, public opinion has started to catch up. Most people now know — and justly criticize — Wilson’s segregationist policies, including the resegregation of parts of the federal government and his infamous White House screening of The Birth of a Nation.

Less known, though, is how much of an absolute disaster he was for free speech. During World War I, Wilson signed the Espionage Act of 1917 and backed the Sedition Act of 1918, which made it a crime to criticize the government, the military, the Constitution, or the war effort in ways officials deemed disloyal. His administration oversaw thousands of prosecutions and arrests, targeting anti-war activists, political dissidents, journalists, and ordinary citizens whose offense was often little more than speaking their minds. Socialist leader Eugene V. Debs was sentenced to ten years in prison for an anti-war speech. Newspapers and magazines were denied mailing privileges for publishing views the administration disliked. Wilson was hard-pressed to meet a dissenter he didn’t want to throw in jail, and in far too many cases he actually succeeded in doing just that.

That’s why FIRE has officially dubbed Woodrow Wilson the worst president for free speech, ever.

Check out this fantastic video from FIRE, which is part of its Figures of Speech series celebrating America’s 250th birthday.