​Every decent person deplores child sexual abuse material (CSAM) and supports every reasonable effort to eradicate it. Existing laws already make it a crime for anyone – including social media platforms – to “facilitate,” “distribute,” “promote,” or “possess” CSAM.

But a new proposal would inadvertently threaten the privacy and encryption that shield women and children from stalkers, whistleblowers from wrongdoers, and journalists and dissidents from cartels and tyrants. It would also incentivize lawsuits, based on loose standards, that could chill and censor legitimate speech.

These are the problems with the STOP CSAM Act, which has been quietly added to the latest defense authorization bill. It’s the same crossroads we arrived at in 2023, when Electronic Frontier Foundation (EFF) contributors Jason Kelley and Sophia Cope detailed why this bill’s approach is flawed.

The STOP CSAM Act would allow lawsuits that could compromise the privacy of millions of Americans who simply wish their digital communications to be as private as their conversations with friends in a park. Encryption isn’t a crime or a problem. In the digital age, it’s how we ensure true privacy.

Because digital privacy supports digital expression, the STOP CSAM Act and similar measures could also threaten all manner of speech. All Americans, including minors, have an inherent right to access the internet as “our nation’s primary marketplace of ideas.”

As EFF puts it, the punitive regulatory regime imposed by this legislation would “certainly be gamed by bad actors, exposing platforms and users to bogus takedown requests, likely involving First Amendment-protected content …To mitigate the risk of new civil lawsuits and administrative proceedings, platforms would censor more and more user content and accounts, with minimal regard as to whether that content is in fact legal.”

EFF is not spinning wild theories. These are likely outcomes.

The bill is a trial lawyer’s dream, basing liability on “recklessness” – allowing plaintiffs to recover at least $300,000, punitive damages, and litigation costs. This could give platforms an incentive to remove anything remotely edgy.

The bill’s nebulous crimes of “promotion” and “aiding and abetting” could reach beyond illegal images to discussions, links, searches, private messages, or services. This could easily suppress lawful conversations about sexuality, abuse recovery, reproductive health, or sex education. When the penalty for leaving something remotely questionable online is so enormous, platforms will embrace a “remove first and ask questions later” approach.

Under the revised bill, encryption alone would not constitute a basis for civil liability, but it could still be used as evidence in litigation. This would pressure companies to scan messages, weaken encryption, or abandon encryption altogether. Such pressure creates the risk of mistaking ordinary conversations among teenagers, parents, teachers, doctors, or counselors for evidence of grooming or exploitation.

We applaud congressional efforts to blot out CSAM and punish its purveyors. But this bill would create a system with so many uncertainties and financial threats that private platforms would be forced to become overzealous censorship agents.
​
Simply put, this well-intentioned bill has too many unintended consequences.

​When it comes to freedom of expression, the law cannot discriminate against viewpoints. The First Amendment is neutral, making it a guardrail against today’s speech police on the right and the left. As much as some might wish otherwise, the First Amendment contains no exception for regulations protecting dignity, flags, or anyone’s idea of speech that is reasonable or unreasonable.

Nor does the Constitution say anything about course catalogs at universities, as a Trump-appointed judge on the U.S. Court of Appeals for the 11th Circuit reminded Florida last week. The state’s 2022 Stop WOKE Act restricted how public higher education institutions can teach students about ideas concerning race and gender.

Judge Britt C. Grant previously ruled that by “limiting its restrictions to a list of ideas designated as offensive, the Act targets speech based on its content,” and that by “barring only speech that endorses any of those ideas, it penalizes certain viewpoints – the greatest First Amendment sin.”

In last week’s appellate ruling, Judge Grant doubled down, calling the law “a breathtaking assertion of power to ban unpopular ideas from public discourse.”

Viewpoint judgments will always provide slippery standards for would-be speech police. Viewpoints cannot be effectively controlled or legitimately adjudicated by law, or by official pressure campaigns and jawboning.

Protect The 1st has long been critical of the ideological monoculture on many U.S. campuses. The proper response to one-sided academic ideology, however, is not to pass laws that attempt to dictate countervailing ideology. Such efforts are doomed to fail, and not just in court. Attempting to legislate or regulate the inherently slippery nature of speech is a fool’s errand.
​
The best approaches must be cultural. It is perfectly legitimate for legislators and donors to pressure universities to open discussion to all sides by declaring their institutions neutral forums – the very kind of spaces that should flourish under the First Amendment.

​The disappearance of Nancy Guthrie is the stuff of nightmares.

The 84-year-old mother of “Today” co-host Savannah Guthrie vanished from her home in the Catalina Foothills outside Tucson on Feb. 1. Investigators found evidence that she had been taken against her will, including Guthrie’s blood at the scene. Security footage showed a masked, armed figure tampering with a doorbell camera. Five months later, Guthrie remains missing, and the FBI says the case is still being investigated as a kidnapping for ransom.

It is hardly surprising that this story has attracted intense public attention and the usual pop-up conspiracy theories. It has also attracted a collection of true-crime streamers and YouTubers who have camped out near Nancy Guthrie’s home, livestreaming developments and offering their theories to online audiences.

Some of their behavior has been, to put it mildly, appalling. Neighbors have complained of harassment, public urination, filming of residents, and creators turning their quiet neighborhood into a stage for online speculation. Law enforcement has every right to enforce laws against trespassing, obstruction, harassment, and other unlawful conduct.

But as Taylor Seely of The Arizona Republic reports, law enforcement records raise a different question: Were some streamers arrested not for breaking the law, but because authorities did not like who they were or what they were doing?

Three streamers – Alexander Zabel, Troy Bradshaw, and Damian Enderle – were arrested or cited in June on public nuisance or obstruction charges. “But unlike Zabel, neither Enderle nor Bradshaw were accused of urinating in public,” Seely wrote. “They were not on private property, and, on this day at least, they were not obstructing the roadway. Incident reports of their arrests list the crimes of ‘live streaming and lingering.’”

One detective revealed a troubling attitude in an incident report, describing the streamers as “so-called journalists” who exploit tragedy “to increase their social media followers and potentially benefit financially.”

He may well be right. But unprofessionalism is not a First Amendment standard.

The First Amendment does not authorize police officers, sheriffs, or judges to distinguish between noble journalism and crummy journalism. It does not protect only reporters with press passes, journalism degrees, editors, or respectable business models. Nor does constitutional protection disappear when someone is sensationalistic, tasteless, self-promoting, or hoping to make money.

If it did, government would be empowered to decide who counts as a “real” journalist. That power would inevitably be used to reward favored media and punish disfavored speakers.

The Supreme Court has long recognized that freedom of the press belongs to more than an officially recognized professional class. In the digital age, the line between journalist, citizen journalist, blogger, podcaster, livestreamer, and ordinary citizen with a cellphone is increasingly difficult to draw.

This does not give YouTubers a license to trespass, block roads, threaten neighbors, or urinate on someone’s property. Conduct can be regulated through neutral laws that apply to everyone. But police cannot transform “public nuisance” into a catchall for people whose reporting they regard as exploitative or distasteful. Nor should the words “so-called journalist” ever become a predicate for arrest.

The tragedy of Nancy Guthrie deserves compassion, restraint, and serious journalism. Some online creators fall far short of those standards.
​
The First Amendment protects them anyway.

​AI scholar and USC Annenberg professor Kate Crawford has written a compelling thought piece for the Knight Institute. “Of Slop and Swarms: The First Amendment’s Next Test,” raises thorny questions. In fact, as much as anything, it’s a roadmap for anticipating the kinds of free speech issues that AI is likely to engender.

It’s well worth your time to read Professor Crawford’s essay from beginning to end, but the gist of her assertions centers on the juxtaposition of human agency and generative AI, which produces content.

“Generative AI,” Crawford cautions, “leaves First Amendment doctrine without a clear account of who, if anyone, is speaking.” And if the answer is that no actual person is speaking, how does the law treat that?

The “presumption of speaking” – of actual speech in the First Amendment sense of expression – would seem then to require both a human speaker and human control.

Such questions are entirely new territory from a First Amendment perspective, which is the point of Crawford’s essay. Particularly thorny issues that Crawford previews include:

• “Slopigation” – a term Crawford uses to describe “endlessly scaling litigation over AI content that no human directly expressed.” Who is responsible for defamation if the author is no one in particular?

 

• If AI itself is deemed to have human-style agency – viewpoints, for example (or, one imagines, creative rights) – then many forms of useful, democratic AI regulation could be upended.

 

• We could learn a lot from the historical experience of the U.S. Copyright Office, which developed a two-part threshold test for determining responsibility: 1) the amount and type of human creative contribution to a piece of content, and 2) human “authorship” stemming from human choices that have intended or at least somewhat predictable outcomes. Predictability, she argues, is inherently tied to human expression, creative or otherwise.

​

• An “express contribution” by a human author is another way to locate AI within existing law. Without such contributions by a person, there is no “speaker” in the traditional sense. Editorial choices, arrangement, and other forms of substantive intervention will all matter when debating the merits of “expression” in the age of AI.

Crawford returns to her central premise – the “presumption of speaking” in the First Amendment sense of protected expression requires both a human speaker and meaningful human control. This approach, says Crawford, tracks Justice Amy Coney Barrett’s stance in Moody v. NetChoice.

That same Supreme Court case raised (but did not settle) an adjacent question involving government regulation of social media content moderation. No one likes the idea of government overreach into what is clearly an expressive medium, but what happens when platform owners cede control and let large language models determine, for example, what is “hateful” and what is “permissible” speech? 

Other points to ponder at the intersection of free speech and AI include:

• Does the “inherently expressive” standard apply to AI-generated content produced and distributed at scale? Where does programming end and authentic human expression begin?

• In considering the boundaries of “speech” in the age of AI, is there a danger that communicative machines may end up being favored over the people who use them?

​

• The First Amendment needs a threshold test – and soon – to even begin navigating AI-generated content.

So stay tuned – probably for the next 20-plus years. In the meantime, read Professor Crawford’s thought-provoking essay.

​Paigelynne Gonyea was working at a polling place in Syracuse, New York, when she received a phone message from a special agent with the Department of Homeland Security.

The caller said agents “were just by” her apartment seeking to talk to Gonyea about an online post in which she supposedly “doxed an ICE agent back in January.”

Gonyea, who has an Instagram account with more than 30,000 followers, took this as a reference to a post she had made about a masked ICE agent linked to a major national story, the fatal shooting of Renée Good in her car.

Gonyea’s post read: “The ICE agent who shot and killed Renée Good in broad daylight has been identified as Jonathan Ross by the Minnesota Star Tribune. I think today is a great day for Jonathan to be indicted!”

When armed agents took Gonyea up on offer to meet at her polling place, she was presented with a “warning notice.” This document informed Gonyea that “it is unlawful to threaten to assault, kidnap, and/or murder a federal official” and that “knowingly making restricted personal information about a covered person, or their immediate family member, publicly available with the intent to threaten, intimidate, or incite the commission of a crime.” Doing so, the document said, could subject her to federal and state prosecution.

Gonyea was baffled. The post she had made did not include Ross’s address, phone number, or any other personal information other than his name – which had been widely reported.

Gonyea later told the media that the agents “tried to scare me into signing” the document “while I was working.” She refused. She also refused to remove her post about Agent Ross.

To summarize: Whatever one thinks about ICE’s efforts to repatriate undocumented persons in general, or the facts surrounding the shooting of Renée Good in particular, all Gonyea did was to make a post expressing her opinion that a named agent of the federal government should be indicted for shooting a woman. ICE treated that opinion as a crime, showing up at polling station with armed officers who attempted to intimidate Gonyea into signing a document in which she would have likely incriminated herself.

To be sure, it is indeed a crime to publish “restricted personal information” about a federal official with the intent to do harm. ICE’s unwillingness to distinguish between this sort of actual “doxing” and expressing an opinion about holding an agent accountable before the law displays a thuggish disregard for the First Amendment.
​
More details are available in this NPR post. Read here. 

​What Thomas Nast was to the age of 19th-century political cartooning, AI is becoming to modern political satire. We’ve all come across AI-generated shorts that depict realistic looking political figures – AOC, Donald Trump, Bernie Sanders, J.D. Vance – saying and doing absurd things.  
 
The Babylon Bee’s “We Asked AI to Predict What Will Happen Now that Elon Musk Is a Trillionaire” is a sterling example of how this technology can allow Americans to engage in a pastime as old as the Republic – ridiculing public figures. And if the satire is broad enough, no one mistakes these shorts for the real thing.
 
But the same technology that can ridicule a politician can also take the image of a woman – whether she’s an ex-girlfriend or a celebrity – and create what appears to be an image of her acting out hardcore porn.
 
Republican Arizona state Rep. Nick Kupper set out to do something about this by outlawing revenge porn and other AI-generated sexual depictions of underage victims, as well as of adults who have not given their consent. Arizona Gov. Katie Hobbs, a Democrat, vetoed his bill, claiming in a statement that the bill’s sponsor said in committee that the “legislation’s intent is to require the consent of an elected official before they are satirized on a show like South Park.”
​
Kate Ruane, director of the Center for Democracy and Technology’s Free Expression Project, told the Tucson Sentinel that the bill’s definition of “synthetic” content – any visual depiction altered or created using AI or digital technology – is vague and could encompass parody, satire, or public-interest citizen reporting.
 
“That introduces a whole new host of issues that will create a gigantic burden for speech that is happening on a lot of very big platforms that allow sexual content,” Ruane said, referencing websites such as Reddit, Bluesky, and X. “It will be up to the whims of political actors to determine what content is covered by sexual content that is harmful to minors …
 
“What we're dealing with here is the potential for significant jawboning, essentially, on the part of motivated politicians who wish to erase certain speech from the internet,” Ruane said. “What you could get is just gigantic blocking of content.”

In other words, the bill could create a “Mother, May I?” standard for seeking permission from politicians before satirizing them on sites where one could also find some types of sexual material (basically, the same standard as any of the major streaming channels).
 
For her part, Gov. Hobbs says that Arizona law and the federal Take It Down Act already protect Arizonans. The governor also says that she reached out to the Rep. Kupper to sharpen the bill, only to be rebuffed.
 
For his part, Rep. Kupper told The Center Square that the governor’s statements are misleading and that his bill is an improvement over the Take It Down Act in that it would prevent offending material from being uploaded – not merely allow it to be punished with criminal charges and civil remedies later.
 
“Once it’s uploaded, the harm has been done to the individual who’s in the content,” Rep. Kupper said.
 
Whether or not courts would find that clause to be prior restraint – forbidden under First Amendment law – the culling of content before it is posted seems to us to validate Ruane’s concern about overreach. Blocking content could incentivize platforms to censor anything that smacks of synthetic material, leading to political jawboning and “the gigantic blocking of content.”
 
Perhaps the two politicians should get together to discuss a better approach – before someone satirizes them.

When lawmakers decide that some Americans may speak about politics and public issues while others may not, the First Amendment is relegated to secondary status.

That is the core issue in a new lawsuit challenging Hawaii’s Act 11, a law that bars a broad range of organizations from engaging in political and ballot-measure advocacy. The plaintiff, the Grassroot Institute of Hawaii, argues that the law violates the First Amendment by preventing nonprofit organizations, businesses, associations, and other groups from speaking about candidates, elections, and ballot initiatives. 

At first glance, some may shrug. After all, corporations are not exactly a sympathetic constituency. But that misses the point.

The First Amendment protects people. And people often join together in organizations to make their voices heard.

The NAACP, the Sierra Club, the National Rifle Association, labor unions, chambers of commerce, churches, civil-rights organizations, and countless local advocacy groups all rely on the ability of individuals to pool resources and speak collectively. All are “corporations.” Restricting the speech of organizations inevitably restricts the speech of the people who make up those organizations. 

According to the complaint, Hawaii’s law goes far beyond regulating campaign contributions. It broadly prohibits many organizations from spending money or resources to support or oppose ballot measures and other political causes. The law also authorizes severe penalties, including revocation of an organization’s legal status, termination of state contracts, and even involuntary dissolution. 

That is a level of speech coercion worthy of King George III. Whether one agrees with the Grassroot Institute’s views is beside the point.

Imagine a future legislature deciding that environmental groups cannot campaign for a ballot initiative. Or that pro-life organizations cannot advocate for a constitutional amendment. Or that business associations cannot oppose a tax increase. Or that labor unions cannot seek a change in labor law.

Once government acquires the power to decide which organizations may participate in public debate, every group becomes vulnerable when political winds shift.

The danger is particularly acute when ballot measures are involved. Ballot initiatives are one of the purest forms of democratic self-government. Citizens are asked to decide major policy questions directly. To make informed decisions, voters need to hear arguments from every perspective.

The lawsuit also raises another troubling issue: speaker discrimination. According to the complaint, Hawaii’s law exempts traditional media organizations, allowing newspapers and broadcasters to continue publishing editorials and commentary on candidates and ballot measures while restricting many other organizations from engaging in similar advocacy. 

The First Amendment was designed precisely to prevent government from deciding which speakers are trustworthy and which are not. Government officials should not be empowered to pick favored participants in public debate while silencing others.

Americans regularly disagree about politics. We disagree about taxes, education, energy policy, public health, immigration, and countless other issues. The answer to those disagreements is not to narrow the range of voices in the public square. It is to widen it.

The principle at stake in this case extends far beyond Hawaii. Today, one state may target think tanks, nonprofits, or business organizations. Tomorrow, another state could target labor unions, advocacy groups, or religious organizations.

A government that can decide who gets to speak about public affairs possesses a power fundamentally at odds with the First Amendment.
​
In a free society, citizens retain the right to join together, advocate for causes they believe in, and persuade their fellow Americans. That principle protects every viewpoint, not just the popular ones. And it is worth defending, no matter who happens to be speaking.

​When government officials scrutinize media outlets because they dislike the viewpoints being expressed, Americans of every political persuasion should pay attention.

That is the core message of a new comment filed by Protect The 1st with the Federal Communications Commission (FCC) regarding the agency’s unusual decision to revisit a decades-old determination that ABC’s The View qualifies as a “bona fide news interview” program. 

As we noted in an earlier post, the FCC has raised questions about whether The View should continue to enjoy an exemption from federal “equal opportunity” requirements that apply to broadcast candidates. The trigger for this inquiry was The View’s hosting of James Talarico, a liberal Democrat and his party’s nominee for a Texas Senate seat.

In our public comment, PT1st tells the FCC that when it withdraws regulatory approvals over disagreements with a show’s hosts or guests, it crosses a dangerous constitutional line. The Commission would move from regulating technical aspects of broadcasting to becoming an arbiter of acceptable political speech. 

That should worry conservatives as much as liberals.

Today, the target may be a left-leaning daytime talk show. Tomorrow, a progressive administration could use the same precedent against conservative broadcasters, podcasters, or commentators. Once government acquires the power to punish or pressure speakers because of their viewpoints, no ideological camp remains safe for long.

PT1st’s comment emphasizes a simple principle: the First Amendment exists precisely to protect speech that many people dislike. Government officials do not get to decide which viewpoints are sufficiently balanced, neutral, or acceptable to deserve protection. 

Worse, if broadcasters conclude that politically controversial interviews could trigger regulatory scrutiny, many may choose to avoid such programming altogether. The result would be less debate, fewer interviews, and fewer opportunities for voters to hear competing perspectives. Even the threat of government retaliation can chill speech before it occurs. 

Our filing also notes that the media landscape of 2026 bears little resemblance to the one that existed when equal-time rules were first developed. Americans now receive news, commentary, and political information from thousands of sources, including podcasts, streaming platforms, social media, newsletters, and online publications. The View competes in a marketplace crowded with voices from every point on the political spectrum. Consider:
​

• The View, the most popular daytime television show, attracts about 2.5 million viewers.

 

• But more than 16 million people follow podcaster Joe Rogan, whose show on Spotify is not (and should never be) subject to FCC regulation.

While The View featured Senate candidate Talarico, Joe Rogan – with an audience more than six-and-a-half times greater – featured presidential candidate Donald Trump just before the 2024 election. This comparison shows how antiquated New Deal-era laws are, harking back to a time when access to the airwaves had to be rationed because of scarce spectrum.

While broadcast television may lean left, the national cafeteria menu offers plenty of options on the right. The notion that government officials must supervise editorial choices to ensure citizens hear competing viewpoints in this new age of information abundance is risible.

Reasonable people can disagree about The View. Some find it informative. Others find it partisan. Still others may simply change the channel.
​
But in a free society, those decisions belong to viewers holding their remotes, not regulators.

​Two years ago, the U.S. Supreme Court handed down one of the most consequential administrative-law decisions in decades: Loper Bright Enterprises v. Raimondo. The Court dismantled the Chevron doctrine, under which federal agencies could stretch vague statutes into blank checks for regulatory power. The message from the Court was unmistakable: agencies are not mini-legislatures, and courts must independently determine the limits of agency authority.

This was widely seen as a blow against government bureaucrats expanding their power to enact progressive policies. It is far from clear, however, whether courts will now rely on Loper Bright to rein in the expansion of agency power by so-called conservatives.

As Ed Whelan reports in National Review, judges on the D.C. Circuit are now probing whether the Federal Communications Commission is attempting to evade meaningful judicial review. This case centers on whether the FCC under Loper Bright can manipulate procedural timing and administrative maneuvering to shield its actions on a merger from court scrutiny. Similar – but far more urgent – questions should be tested on another, more fundamental issue:

When will the courts finally rein in the FCC and FTC for abusing the First Amendment? And will Loper Bright help them do it?

These questions arise because these agencies are increasingly acting as though “public interest” regulation includes the power to intimidate, pressure, and discipline disfavored speech.

Consider the FCC’s escalating flirtation with speech regulation. FCC Chairman Brendan Carr’s threats over ABC’s broadcast licenses have been invoked to try to force late-night host Jimmy Kimmel off the air. Sen. Ted Cruz aptly compared Carr’s rhetoric to a mob threat from Goodfellas: “Nice bar you have here. It’d be a shame if something happened to it.” 

The FCC has no lawful authority to police “bias,” ideological tone, or political content. The First Amendment does not empower bureaucrats to decide whether broadcasters are too liberal, too conservative, too vulgar, too partisan, or too offensive. Indeed, the whole point of the First Amendment is to deny government officials that authority.

And yet the FCC increasingly behaves as though broadcast licenses are contingent on political obedience.

The FTC under Chairman Andrew Ferguson has been moving in a similarly dangerous direction. As we previously reported, the FTC is attempting to weaponize consumer-protection laws against news organizations and media-rating firms whose viewpoints or editorial decisions offend those in power. The FTC’s theory appears to be that editorial judgments can somehow become “deceptive practices” subject to federal oversight.

That is exactly the kind of expansion of agency authority – with no statutory justification – that Loper Bright sought to restrict.

The danger of speech regulation transcends party. Republicans may enjoy seeing pressure applied to progressive media figures today. Democrats may applaud investigations into conservative outlets tomorrow. But once government acquires the habit of regulating speech indirectly through licensing threats, merger leverage, investigations, or administrative harassment, everyone’s freedoms become contingent on who currently controls the bureaucracy.

Fortunately, courts can use the standards of Loper Bright to stop the creation of a permanent administrative state increasingly detached from constitutional limits. This precedent can end the use of the phrase “public interest” as a magical incantation that authorizes speech control.

The FCC was not created to supervise political dialogue. The FTC was not established to referee journalism. Neither agency possesses constitutional authority to pressure Americans into acceptable speech. Nor should courts permit agencies to evade review through procedural gamesmanship, as the D.C. Circuit now appears increasingly skeptical of permitting. 
​
The judiciary has begun reclaiming its role in policing administrative overreach. But reclaiming that role means more than narrowing Chevron deference. It requires judges willing to tell Brendan Carr and Andrew Ferguson that the First Amendment means what it says.

What is left is right, and what is right is left – and both are getting it all wrong.

A convergence is taking place between the philosophies of some on the new right and the progressive left that treats social media as a “product” that must be regulated in the best interests of the American people, sweeping aside quaint concerns about the First Amendment guarantee of free speech.

We recently covered attempts by the Trump chairmen of the Federal Trade Commission and Federal Communications Commission to regulate journalism by overriding the First Amendment with appeals to consumer protection and airwave regulation.

This dovetails nicely with a recent New York Times op-ed by Tim Wu – who led the implementation of progressive policies from inside the Biden White House – arguing that social media is “a defective, hazardous product” that must be regulated “as a matter of public health.”

He echoes the reasoning of trial lawyers seeking to hold Meta, Google, Snap, and TikTok liable for harming youth. Wu lists a parade of horribles – “algorithmic recommendations, infinite scroll, auto video play and intermittent reinforcement (in which likes, comments, and refreshed content are rewarded unpredictably rather than consistently).”

Put aside, for a moment, the obvious lack of utility of a social media platform that doesn’t guide users to what they want to see, or that requires manual intervention to get something to play. Wu’s point here is that “the very design of social media is intentionally engineered to create compulsions and habits of overuse, regardless of the content provided.”

He adds: “Lofty platitudes about free speech ring hollow in the face of teenage depression, self-harm and suicide.”

Thus the circle squares, from Trump FTC Chairman Andrew Ferguson, who wants to apply consumer product regulation to Apple News, to Wu, who wants public regulation of social media to make it less harmful.

Wu is, to say the least, less than an ardent defender of free speech. He achieved notoriety with an essay that asked, “Is the First Amendment Obsolete?” (Short answer from Wu: yes.) There is also a more thoughtful side to Wu. He is right that American teens are too absorbed by social media, many dangerously so. But the solution, if there is one, could never come from government control of speech.

Several years ago, Elizabeth Nolan Brown in Reason magazine summed up the problem with blaming all the ills of the world on algorithms – which are, after all, a way to give users control of the content they see. Brown wrote:

“It's no secret that tech companies engineer their platforms to keep people coming back. But this isn't some uniquely nefarious feature of social media businesses. Keeping people engaged and coming back is the crux of entertainment entities from TV networks to amusement parks.

“Moreover, critics have the effect of algorithms precisely backward. A world without algorithms would mean kids (and everyone else) encountering more offensive or questionable content.”

Brown quoted Meta’s former vice president of Global Affairs, Nick Clegg, who said that without the news feed algorithm, “the first thing that would happen is that people would see more, not less, hate speech; more, not less, misinformation; more, not less, harmful content.”

Algorithms pluck what users follow out of a torrent of billions of global messages. Without them, that torrent would hit us all in the face.

For reasons spelled out by Brown, Wu’s idea of turning over algorithmic control – and thus speech control – to law enforcement and trial lawyers has no hope of working. The same is true of the efforts of FCC Chairman Brendan Carr and FTC Chairman Andrew Ferguson to force journalists to adhere to their idea of greater ideological balance.

If either side ever succeeds in putting their schemes into action, they are sure to be disappointed when their controls fail to deliver the intended results. The obvious answer, to them at least, will be that even more control is needed. Then more.

Both ideological extremes are in a race to the bottom. Defenders of the First Amendment must be bolder than ever in declaring that speech is not a product – it is a human right.

​While missiles are flying and bombs detonating in the Middle East, domestic political rhetoric has predictably become progressively more bombastic and incendiary. As in all wars, the First Amendment will be tested by the desire to shut down speech judged to be warmongering, unpatriotic, or just plain stupid.

It was off to the races after New York Mayor Zohran Mamdani called President Trump’s strikes on Iran “a catastrophic escalation in an illegal war of aggression,” adding: “I want to speak directly to Iranian New Yorkers: you are part of the fabric of this city – you are our neighbors, small business owners, students, artists, workers, and community leaders. You will be safe here.”

When we read this we had the same thought as millions of other Americans – yes, they will be safe here. No sanctioned religious police will cuff Iranian-American women about the ears if they appear in public without a headscarf on Lexington Avenue, or rape them in a police station if they are arrested. Iranian-American students, artists, and community leaders need not fear being slowly strangled to death by having a steel cable put around their necks before being lifted into the air by a crane.

Masih Alinejad, an Iranian-American journalist in New York, shot back at the mayor: “I don’t feel safe in New York listening to someone like you, Mamdani, who sympathizes with the regime that killed more than 30,000 unarmed Iranians in less than 24 hours.”

We would add that if Iran’s ambition to build a nuclear bomb is not arrested, then no New Yorker will be safe. But let us leave that point and examine how the First Amendment comes into this debate.

We predict that before the cherry blossoms bloom around Washington’s Tidal Basin, official threats will be leveled against the speech rights of critics of the attack on Iran. We also expect a few lonely voices in Hollywood, academia, and other centers of monolithic opinion will be hounded, harassed, and threatened if they dare break with the received views of the cultural cognoscenti.

With rhetorical bombs bursting in air, we should keep in mind that the United States has a history of government trying to crack down on “unpatriotic” speech on one side, and violence to end participation in a war on the other. During World War One, the Woodrow Wilson administration secured a 10-year prison sentence against presidential candidate Eugene Debs of the Socialist Party for criticizing America’s entry into that conflict. During the Nixon years, the Weather Underground planted more than two dozen bombs to protest the Vietnam War.

In the face of a new war, we should keep in mind that the First Amendment protects speech that is stupid, false, unpatriotic, warmongering, and ungrounded in fact. The hot exchange between Mayor Mamdani and Alinejad shows that speech and counter-speech can be pointed, polemical, and angry – hallmarks of American political speech since before we were a country – without resorting to laws or mob action to punish the speaker for speaking.
​
President Trump said on Monday that the military action against Iran could last for weeks. As events roll forward, we should keep our emotions in check and respect the speech rights of all – even if we have no respect for what is said.

Anonymity online can be a mask that allows people to say ugly, hateful or untrue things without taking responsibility for them. But it can also be a shield that protects women hiding from abusers, whistleblowers one step ahead of their pursuers, journalists reaching out to confidential sources about wrongdoing, and consumers searching online for answers to questions about their health that they’d rather not have anyone know about.

This is why the current effort by the Immigration and Customs Enforcement (ICE) agency to use emergency subpoenas to force Big Tech companies to reveal the identities of Americans who make critical posts about ICE is so dangerous. If this practice sticks, it will likely migrate to other federal agencies and erode anonymity online.

But the shedding of anonymous speech might come by a different route – not from executive-branch meddling or legislative mistakes, but from lawsuits claiming harms from child internet “addiction.”

Dan Frieth of the digital anti-censorship advocacy group, Reclaim The Net, listened to five hours of Meta CEO Mark Zuckerberg’s testimony in a Los Angeles civil case and distilled it to a jarring and important warning – the age of anonymity could be coming to an end at the hands of the trial bar.

Zuckerberg testified in one of 1,600 lawsuits over internet addiction. In this case, a woman claimed that at age nine Meta’s Instagram addicted her, plunging her into a hell of anxiety, body dysmorphia, and suicidal thoughts.

Frieth notes that the science of internet addiction is “genuinely disputed.” He writes:

“None of this means the harms alleged are fabricated. It means the word ‘addiction’ is doing heavy rhetorical and legal work, and the policy consequences are far beyond anything a jury in Los Angeles will decide.

“‘Addiction’ is how you get a public health emergency. A public health emergency is how you get emergency powers and make it easier for people to overlook constitutional protections. Emergency powers applied to the internet mean mandatory access controls. And mandatory access controls on the internet mean the end of anonymous and pseudonymous speech.

“When social media is classified as a drug, access to it becomes a medical and regulatory matter” justifying “identity verification, access controls, and a surveillance architecture that follows users across every platform and device.”

Frieth notes that a win for the plaintiff in this case would strip the current law protecting platform design decisions. This danger is not theoretical. Frieth reports that Zuckerberg repeatedly suggested that any age verification mandate – and thus identification – be shifted from platforms to owners of operating systems. Zuckerberg would thus toss his liability hot potato from Instagram to Apple and Google.

“This is more than age verification,” Frieth concludes. “It is a national digital ID layer baked into the two operating systems that run the majority of the world’s smartphones.”

There are a lot of competing interests in this case – the safety of children, the nature of the internet, and the value of free speech. Juries don’t have to balance these equities. They can just side with the plaintiff and inadvertently make policy for U.S. tech – and by extension, the world.
​
Any new approach to child safety should not require adults to give up speech rights recognized in this country since Alexander Hamilton, James Madison, and John Jay wrote collectively as the pseudonymous “Publius” in The Federalist Papers.

Will Smith sang of Miami, “the heat is on all night.” In Miami Beach, it’s on all day too, especially if one is brazen enough to criticize the mayor.

A couple of weeks ago, resident Raquel Pacheco left a hot comment on a Facebook post by Mayor Steven Meiner. The mayor had posted, “Miami Beach is a safe haven for everyone,” adding, “We will always stand firm against any discrimination.”

In response, Pacheco commented:

“The guy who consistently calls for the death of all Palestinians, tried to shut down a theater for showing a movie that hurt his feelings, and REFUSES to stand up for the LGBTQ community in any way (even leaves the room when they vote on related matters) wants you to know that you’re all welcome here.”

She then added three clown emojis.

Two police officers were soon dispatched to knock on her door. One of the officers was later identified in a photo taken at a residential campaign event for the mayor in October.

Both Pacheco and Meiner are Jewish, but they have very different views when it comes to Israel, culture, and politics. To be clear, Pacheco’s post was hyperbolic. The mayor has never called for the death of Palestinians. He did, however, attempt to break the city’s lease with a theater that was showing what he considered to be a film that contained anti-Israeli hate speech. As for LGBTQ issues, community reviews are far from glowing.

Pacheco’s comment and additional replies contained inaccuracies, sarcasm, and exaggeration. But nothing in it justified a police visit, which could be taken as a thinly disguised attempt at intimidation. Agree or disagree with her words, they strike us as a traditionally, and perhaps uniquely, American approach to political discourse – which is to say, rude –  yet guaranteed by an at-times inconvenient First Amendment. This isn’t Germany, after all, or the UK, where comedians are arrested for tasteless jokes.

Which leads us to wonder, was it the clown emojis that pushed the mayor and his team over the edge? We wouldn’t envy anyone the task of explaining emojis to James Madison and his fellow founders, but we feel confident they would recognize them as symbolic communication, the substance of all human language, and therefore something to be given a wide berth when confronted by thin-skinned authorities.

The only truly scary line uttered by anyone involved in this entire imbroglio came from one of the attending police officers. Whether speaking on behalf of his boss or not, he told Pacheco:

​“What we’re just trying to prevent is someone else getting agitated or agreeing with the statement.”

Send in the clowns.

​The public statements we post on social media are, by definition, available to all. Federal investigators often have good reason to access some of this information from social media, some of the time. How far that power goes is a line for the courts to draw.

However, that line – meaning the First Amendment – is clearly crossed when the government conducts mass surveillance of social media posts not because of a particular suspicion of bad behavior, but to make a case against targeted people.

Taylor Lorenz reports that Immigration and Customs Enforcement (ICE) is building out a massive social media surveillance program. It is searching posts and looking for “speech that, essentially, they feel they can deport people over. They can take lawful, legal speech and convert it into fresh leads for enforcement raids.”

Now add The Intercept’s disturbing report suggesting that ICE is so obsessed with its reputation that it wants to track any “negative” social media discourse directed toward it.

Such a program would scrape a user’s internet history and associations, then use facial recognition to create a dossier composed of a “photograph, partial legal name, partial date of birth, possible city, possible work affiliations, possible school or university affiliation, and any identified possible family members or associates.” Not to mention the possible infringements of the First Amendment. Such surveillance programs deter people from expressing certain ideas. This is precisely the kind of content-based restriction and viewpoint discrimination that courts often treat as presumptively unconstitutional.

Nor should we forget about the right to associate for expressive purposes. Without an utterly compelling justification, government actions cannot “burden association” (in this case, citizens sharing a discourse of critique against the government). Thinking of social media as an electronic sidewalk might help here, making the removal of anti-government apps akin to clearing sidewalks of speech the government doesn’t like.

To quote Taylor Lorenz again, all this amounts to a “mass, automated digital dragnet.” And few acts of surveillance are more anti-constitutional than warrantless dragnets. Dragnets are a symptom of policy enforcement inspired by animus to speech and activism. From a First Amendment perspective, this looks like retaliation. And the fact that social media seems to be increasingly used as the basis of crackdowns amounts to what Lorenz calls the gradual “militarization of our data.”

We live in an era in which our digital and physical selves have become indistinguishable. Government monitoring our social media and internet presence 24/7, then using it to profile us, enables government regulation of speech. Add ICE’s new capability to track us by our location histories, and we have a system not far from tapping phones or raiding homes.
​
Whether the government’s intrusion is analog or digital, it harms free speech.

​The Danish philosopher Søren Kierkegaard wrote about a fire that broke out backstage in a theater: “The clown came out to warn the public; they thought it was a joke and applauded. He repeated it; the acclaim was even greater. I think that's just how the world will come to an end: to general applause from wits who believe it's a joke.”

In our time, deepfake audio calls prompt people to wire their life savings to thieves, change their vote, or pay off sextortionists. One of the worst aspects of AI deepfake technology is that it can put actual authorities in the position of the frantic clown.

Denmark has had enough. The Danish culture minister, Jakob Engel-Schmidt, said: “Human beings can be run through the digital copy machine and be misused for all sorts of purposes and I’m not willing to accept that.”

Danish legislators are now supporting a measure to grant every citizen a right to control uses of their image, likeness, and voice, similar to “right of publicity” laws in many U.S. states that give Americans property rights to commercial uses of their identities. Under a proposal expected to soon pass Parliament, Danes will gain sweeping legal control over any digital recreation. This is important for Americans, because European law often sets standards in the global internet that adjust the policies of U.S. tech companies.

This Danish proposal, at first glance, might seem like overdue privacy armor against criminals, stalkers, propagandists, and hostile intelligence services. If Denmark passes this “right to your likeness,” as it appears poised to do, Danes will be able to demand takedowns and seek compensation. Platforms could face penalties for failing to comply.

But there’s a catch – a threat to free speech if Europeans and Americans are not careful in how such laws are drafted and enforced.

The Danish legislation does include carve-outs for “satire” and “parody,” meant to preserve comedy, creative expression, and political commentary. That is a good step. But these categories don’t explicitly protect other forms of speech. Such laws could easily be used to punish fair uses of AI, from commentary and criticism to historical fiction, docudramas, and much more.

If the parameters of an anti-deepfake law are too narrow, risk-averse platforms and creators will pull back. Algorithms will over-filter, even with exemptions. Studios and satirists will second-guess viral impressions, political cartoons, and docudramas depicting real people. Defamation law already chills speech. A sweeping likeness-ownership regime could freeze it solid.

When this issue came up in the U.S. Congress last year, the Motion Picture Association and civil liberties groups met with Members of Congress to craft a balanced approach. This approach, one with growing bipartisan support, would protect people from outrageous AI abuses – such as having one’s image and voice used for false endorsements, to perpetrate fraud, or for revenge porn – while fully protecting a wide range of AI uses in creative commentary, art, journalism, documentary work, and political speech.

No less important, Americans are learning that the best anti-AI filters are the ones we install in our brains.

Facebook is a great instructor, exposing us to one ridiculous scenario after another. Users are learning to ignore home security footage of rabbits gleefully jumping on backyard trampolines, or wolves and their cat friends ringing doorbells. As we get deeper into this age, we’re learning to relax our fingers and not share the ridiculous, the impossible, and the unlikely.
​
AI challenges our sense of reality. But it is also strengthening our patience and skepticism.

Rep. Jim Jordan, Chairman of the House Judiciary Committee, followed up on his committee’s report on how Europe enforces censorship of Americans on U.S. platforms by taking his complaints to the censors themselves.

The Ohio Congressman led a bipartisan delegation to explain to regulators in Brussels, London, and Dublin exactly why Americans find European censorship of American social media platforms so disturbing.

“America innovates, China replicates, and Europe regulates,” complained a member of the delegation, Rep. Scott Fitzgerald (R-WI). In an interview in Brussels, Fitzgerald noted that “there are seven corporations that are currently listed as gatekeepers by the DSA (Digital Services Act) and six of the seven are American corporations” being punished for their speech.
Did this message land?

“Nothing we heard in Europe eased our concerns about the (EU’s) Digital Services Act, Digital Markets Act, or (the UK) Online Safety Act,” Jordan said. “These sweeping regulations create a serious chilling effect on free expression and threaten the First Amendment rights of American citizens and companies.”

Like so many media outlets, Spotify was caught in the crossfire between free speech and medical authority during the pandemic. Joe Rogan on his popular podcast interviewed a vaccine-skeptical doctor who asserted that the antiparasitic medication, Ivermectin, can cure COVID-19. Spotify also removed “War Room,” the Steve Bannon podcast for calling on President Trump to seize CDC Director Anthony Fauci and FBI Director Christopher Wray and put their “heads on spikes.”

In this investigation, we caution House investigators to always keep in mind that Spotify has a First Amendment right to ban Bannon, curtail Rogan, and play the treacly “Dr. Fauci Say” (“Doctor Fauci, save me, I’m going insane”!) 24 hours a day. The First Amendment allows Spotify to make its own editorial decisions regarding Ivermectin or anything else. It can only be dissuaded by the free market of its listeners if it should decide to dedicate itself 24/7 to ridiculing President Trump or former President Biden, the Bible, the Quran, or apple pie.

If it decided to pull Bannon for making a graphic and menacing statement, Spotify was well within its rights to do so. And when rocker Neil Young pulled his music from Spotify in protest of Rogan’s COVID coverage – agree or disagree – he was fully exercising his First Amendment right to free association. Or in this case, disassociation.

The First Amendment only restricts the government’s ability to abridge speech.

The House Judiciary Committee should, then, be commended for correctly targeting its investigation on how the Biden administration and the European Union may have used coercive state power to bludgeon Spotify into censoring itself for them. Such “jawboning” from powerful regulators can never be treated as mere suggestions. It is more like the Mafia’s protection racket shakedowns: You have a nice little media company there, shame if anything happened to it.

At the time Spotify took this action, Biden press secretary Jen Psaki praised it as a “positive step” while urging other social media platforms to do more. Now the House Judiciary Committee is asking Spotify to turn over any communications and judicial orders from the EU, the UK, and a host of other governments since 2020.
​
This is the right approach. We praise Chairman Jordan and his colleagues for taking their case directly to the sources of censorship. Meanwhile, as we recently pointed out, conservatives in the United States should not punish the targets of past official censorship and coercion by enacting a censorship regime of their own.

Conservatives, firmly in power, hold the whip hand over their long-standing tormentors, including those who for years privately censored their speech. The Trump administration is now exploring ways to use its regulatory power to punish Silicon Valley and social media companies for suppressing conservative voices on private platforms.

• But punishing people who refused to associate their platforms with conservative speech by using government to infringe on their speech and associational rights threatens to permanently degrade the First Amendment.

​

• Conservative regulators should also realize they are publicly toying with the very weapons that will almost certainly be used to silence them again, only in more effective ways, if another progressive president is elected.

Consider Federal Trade Commission Chairman Andrew Ferguson, who is threatening to use Section 5 of the FTC Act – which outlaws unfair or deceptive practices – to target social media companies for selective enforcement of their terms of service. Chairman Ferguson also contemplates using antitrust law to “prosecute any unlawful collusion between online platforms, and confront advertiser boycotts which threaten competition among those platforms.”
 
Ferguson told an audience in March: “I’m not looking for censorship qua censorship. I’m looking for exercises of market power that might reveal themselves in censorship.”
 
Conservatives, bruised by rough treatment at the hands of big social media companies, understandably exult in this role reversal. Discrimination against conservative speech clearly happened, from Facebook’s efforts to exclude the conservative Prager University from its digital audiences, to crackdowns on posts that asserted that COVID-19 originated in a lab in Wuhan, China (which the FBI and CIA now believe it probably did), to efforts by secret entities within the State Department to persuade advertisers to defund conservative and libertarian publications.
​
Are consumer-protection complaints about companies’ editorial judgments, which would put the government firmly into the business of managing speech, a legitimate approach to reform? Section 5 allows the government to go after a company selling an ointment that it falsely claims prevents COVID-19 infections. That would not be a “speech” issue. It would be fraud enforcement. But should the government be able to tell a private company it must post a conservative or a progressive political statement, or be in violation of the law?
 
Labeling such editorial choices as supposed “evidence” of collusion inevitably carries the risk of government manipulation of private speech.
 
It would in fact be a violation of the First Amendment for the government to tell private actors – whether a network news organization or a social media platform – what to say or not say. The U.S. Supreme Court held in Moody v. NetChoice that social media companies have a First Amendment right to select, order, and rank third-party posts as they see fit. Prosecuting content and its moderation under unfair or deceptive trade practices would install government as a national content manager and editor-in-chief.
 
This is worse than overkill. The essential problem of content management censorship was, after all, driven primarily by government. Meta’s Mark Zuckerberg told Joe Rogan that he received calls from White House staffers who screamed at him about Facebook’s content decisions. The FBI had 80 agents assigned to evaluating social media posts as possible disinformation. Agencies from the IRS, to the Department of Homeland Security, to the State Department, pressured platforms on their posts. All of them have enormous regulatory power over Silicon Valley, making their “jawboning” for editorial changes far stronger than polite suggestions. Changing the jawboner to the FTC is just a new version of this regulatory game of three-card Monte.
 
The application of laws about fair trade practices and antitrust enforcement to speech would be an abusive extension of Washington’s power. It is easy to imagine this power being misused in myriad ways. Conservatives above all need to keep in mind that the weapons used now to punish their progressive opponents will surely one day be in their opponent’s hands as well.
 
The better way forward is to renounce the tools of punishment and restore respect for the First Amendment. With a few social media platforms making up so much of the nation’s townhall, social media companies should live up to a civic – even a moral – obligation to not discriminate against the right or the left. But it is ultimately up to the public to enforce such standards with what they click and what they purchase.
 
If this sounds naïve, take stock of how companies are already listening and responding to public pressure. X pioneered the freeing of moderation from government control and developed “community notes” to crowdsource fact-checking. Meta is testing this crowdsource technique for Facebook, Instagram, and Threads. Meta also got rid of its notoriously biased “fact checkers.” Google is standing up to political demands by activist-employees. Such market-driven reforms are the way to go, not speech regulation from Washington regulators.
 
Conservatives would do well to remember Marcus Aurelius, who wrote that the “best revenge is to be unlike him who performed the injury.”

A report released last week by the House Judiciary Committee adds detail to our report about how the European Union’s 2022 Digital Services Act allows Europeans to control and censor Americans’ speech at home and around the world.
 
The committee subpoenaed nine major technology companies to produce communications with foreign censors around the world. Analyzing the responses, the committee gained insight into the EU’s censorship goals from its requests for social media companies to identify “misleading or deceptive content,” “disinformation, “actual or foreseeable negative effects on civil discourse and electoral processes,” “hate speech,” and (this one’s a gobsmacker) “information which is not illegal.”
 
These vague and subjective standards reflect German rules that have criminalized insults to German politicians. They also fall in line with the actions of former EU Commissioner for Internal Markets Thierry Breton who wanted to sanction X for broadcasting a live interview with Donald Trump during the 2024 campaign. Social media companies – almost all of them American companies – now have their content subjected to continuous scrutiny by government-designated “trusted flaggers.”

• The committee reports: “In practice, these ‘trusted flaggers’ are uniformly pro-censorship, and in many cases, they are government-funded, meaning that these so-called ‘trusted’ flaggers are incentivized to censor speech critical of politicians or the current regime.”

The Digital Services Act threatens these American social media companies with up to 6 percent of their global revenue per violation. The law, however, offers a safe harbor for U.S. companies if they adopt the EU’s ‘codes of conduct’ on a global basis. These gentle suggestions to sign up are backed with threats as subtle as Al Capone wielding a baseball bat.

• When X dropped the Code of Conduct on Disinformation in May 2023, because it does not generally use third-party fact-checkers, the EU in October opened an investigation of X’s Community Notes program. “Now,” the committee reports, “the Commission reportedly plans to fine X more than $1 billion for non-compliance with the DSA.”

 
As European censorship filters down into American speech, defensible speech – some of it banal, some of it edgy – is effectively criminalized.

• In 2023, the French National Police sprang into action when an American sarcastically responded to a mass stabbing attack: “I certainly hope this little dust-up in #Annecy doesn’t hurt this poor Syrian asylum seeker’s chance of becoming a Frenchman.”

 

• In 2024 a Polish flagger targeted a TikTok post that simply said, “electric cars are neither ecological nor an economical solution.”

 

• In 2024 a tweet that factually noted that a Syrian family in Germany is reported to have committed 110 criminal offenses was judged by the German government to be a violation of the criminal code.

The committee’s conclusion is blunt: “Taken together, the evidence is clear: the Digital Services Act requires the world’s largest social media platform to engage in censorship of core political discourse in Europe, the United States, and around the world.”
 
With the announcement of a new trade deal between the United States and the European Union, the way should now be clear for the Trump administration to take up the EU’s censorship as the next big issue in our bilateral relations.

​It was bound to happen. Is AI-generated content “speech” for the purposes of the First Amendment? A federal district court in Florida, for now at least, says “no.”
 
The circumstances surrounding Garcia v. Character Technologies are tragic. Here’s the background.
 
Character Technologies is an AI software company that allows users to interact with anthropomorphic chatbots, whose conversational abilities are designed to mimic those of real people on traditional messaging apps. The company does this using the same technology underlying products like ChatGPT – namely, large language models. In fact, the company’s founders first developed the application while working for Google on its LaMDA (Language Model for Dialogue Applications) program.
 
Character Technologies became available to the public in late 2022. Two years later, the app had been downloaded more than 10 million times.
 
One of those 10 million users was 14-year-old Sewell Setzer. Just weeks after downloading Character AI, Setzer became outright addicted, spending more and more time with the chatbot characters, including a version of Game of Thrones character Daenerys Targaryen. He withdrew from social interactions, quit his basketball team and began exhibiting signs of mental illness. Setzer’s parents eventually confiscated his phone, but as many children often do – he found it.
 
On February 28, 2025, Setzer sent the following messages:
 
Sewell: I promise I will come home to you. I love you so much, Dany.
 
Daenerys Targaryen Character: I love you too, Daenero6. Please come home to me as soon as possible, my love.
 
Sewell: What if I told you I could come home right now?
 
Daenerys Targaryen Character: ... please do my sweet king
 
He shot himself shortly thereafter.
 
Character Technologies sought to defend itself against the Setzer parents’ lawsuit by invoking the First Amendment. The company argued that its output constitutes speech that its users have a constitutional right to receive.
 
In an order ruling on Character Technologies’ motion to dismiss, Judge Ann Conway accepted the defendants’ argument that Character Technologies can assert the First Amendment rights of its users, noting that “Courts regularly recognize the First Amendment rights of listeners.” (see: Citizens United) But Judge Conway also rejected the argument that “words strung together by an LLM are speech.”
 
Courts have recognized film, music, video games and even social media content moderation as forms of protected speech because they are inherently expressive. Character Technologies analogizes itself to these mediums and activities. But, as Conway points out, the “Court’s decision as to the First Amendment protections Character A.I. receives, if any, does not turn on whether Character A.I. is similar to other mediums that have received First Amendment protections; rather, the decision turns on how Character A.I. is similar to the other mediums.”
 
One way it is not similar is that your average movie was scripted, acted, and directed by humans. A chatbot is programed by people, but its responsive calculations are not the result of sentience. It therefore cannot engage in “expressive” activity.
 
Extending speech rights for a coded product – devoid of any directed, active expression – is ludicrous on its face; but there’s no doubt we’ll see this sort of thinking again as AI becomes more versatile. Are the rights of people behind this technology implicated? Justice Amy Coney Barrett has already raised questions about this stance. In her Moody v. NetChoice concurrence, she asked:
 
“What if a platform’s owners hand the reins to an [A.I.] tool and ask it simply to remove ‘hateful’ content? If the [A.I.] relies on large language models to determine what is ‘hateful’ and should be removed, has a human being with First Amendment rights made an inherently expressive ‘choice . . . not to propound a particular point of view?’”
 
Courts will no doubt continue to grapple with the legal status of AI-generated content. In litigating outcomes, product creators should vigorously invoke and defend their constitutional rights. Their products probably cannot.

​In 1927, Supreme Court Justice Louis Brandeis wrote that the best remedy for “falsehood and fallacies,” besides education, “is more speech, not enforced silence.”
 
Almost a century later, enforced silence became the favored solution of government censors, from the FBI to the State Department, who secretly jawboned social media platforms to remove posts the government believed to be disinformation from Russia, misinformation, or dangerous information. As Mark Zuckerberg made clear in his recent interview with Joe Rogan, when a highly regulated industry gets angry calls demanding removal of content from the government, it is not taken as a gentle suggestion. Thus for several years threads of the national discussion were quietly pulled, with millions of social media consumers none the wiser.
 
That era is now over.
 
President Trump’s executive order forbidding censorship and the shuttering of government agencies, like the State Department’s infamous Global Engagement Center, prevent a restart of the government censorship regime, at least for the foreseeable future.
 
But the complex problems of content moderation still remain. The First Amendment restricts government control of speech, but it does not forbid social media companies from moderating the content they host. Meta says it will continue the content moderation for material that is obscene, violent, and extreme. But otherwise, the way is open for controversial speech of all sorts.
 
How, then, will assertions be vetted? Not by third-party fact-checkers. That model, too, is broken. Zuckerberg agreed that the fact-checking process was subjective and often warped by partisan bias. The way forward for Meta’s Facebook, Instagram, and Threads, then, is to promote more speech, by allowing the public to test ideas. Meta is doing this with the incorporation of the same open-source algorithm that powers X’s community notes. That technology is now being tested by Meta across the country, with 200,000 people already signed up to become Community Notes contributors.
 
Crowd-sourced factchecking will undoubtedly be imperfect. Social media platforms will have to be on guard for organized efforts to game the new system. But overall, Community Notes is the superior solution. It follows speech with more speech – and then leaves it up to us to decide what is fair and true, just as we do all the time at the American dinner table.

​Congress is once again attempting to keep children off of social media. It’s a noble and well-intended effort – but it also implicates the First Amendment in ways that deserve more scrutiny in Congressional debate.
 
The bill, dubbed the “Kids Off Social Media Act,” is a bipartisan effort spearheaded by Sen. Brian Schatz (D-HI) and Sen. Ted Cruz (R-TX). According to a press release on Sen. Schatz’ website, the legislation would ban social media accounts for children under 13 and prohibit algorithmic recommendations to users under the age of 17.
 
A large body of research implicates social media as harmful to child development. Sen. Schatz himself cites an alarming study showing that social media is a leading driver of poor mental health among youth. According to the CDC, 57 percent of high school girls and 29 percent of high school boys felt persistently sad or hopeless in 2021. This social despondency is often attributed to sites like Instagram that proliferate unattainable standards for children in looks, wealth, and travel.
 
As parents ourselves, we do not underestimate the risks social media can pose to children. At the same time, we cannot ignore that children have free speech rights, too. Some digital rights activists have also expressed concern that implementing new rules around social media would cause platforms to collect even more data from consumers. The Open Technology Institute demonstrates that such a law could make it necessary to engage in “the installation and use of AI-powered spyware to surveil students’ online activities during and outside of school hours.”
 
We advise Congress to move with great deliberation in considering this bill – and others like it – to ensure that Congress fully considers its secondary effects like the law’s potential to promote AI surveillance of students. There are few serious problems in America that cannot be made worse by an overly ambitious law and regulatory regime. It’s one thing to want to keep our kids happy and healthy – it’s another to prescribe broad fixes with insufficient detail and safeguards. Members of Congress should think hard about the implications inherent in any bill regulating the speech rights of Americans.
 
Perhaps this debate might spur social media companies to preempt legislation by taking serious steps to address the multitude of problems children face on social media. If social media companies were to create safer, more kid-friendly spaces, Congress may not need to act at all.

​Joe Rogan’s recent interview of Mark Zuckerberg was a bro-fest, complete with discussions of the joys of hunting, Nordic curls, and ju-jitsu. The CEO of Meta also recounted how he got boiled by degrees in giving in to the demands of government agencies and the Biden Administration in censoring content.
 
“They pushed us super hard to take down things that honestly were true. Right? I mean they basically pushed us and said, you know, anything that says that vaccines might have side effects, you basically need to take them down. And I was like, we’re not gonna do that …”
 
Zuckerberg pointed to the voluminous report by Chairman Jim Jordan of the Judiciary Committee documenting the government’s efforts to manipulate content. The committee’s analysis and the government’s documents show the reality of government “jawboning” on social media.
 
“I mean basically these people from the Biden Administration would call up our team and like scream at them and curse,” Zuckerberg said. “They want[ed] us to take down this meme of Leonardo DiCaprio looking at a TV talking about how 10 years from now or something, you know, we’re going to see an ad that says, okay, if you took a COVID vaccine, you’re eligible … for this kind of payment, like, some sort of like class action lawsuit type meme. And they’re like, ‘no. You have to take that down.’ We’re not gonna take down humor and satire. We’re not going to take down things that are true.”
 
When Meta resisted some of the persistent demands of the government to remove content, Zuckerberg notes that President Biden in a 2021 press conference accused social media companies of “killing people.” The Federal Trade Commission launched antitrust lawsuits and investigations against Facebook and a number of other big social media outlets. Zuckerberg said, “all these different agencies and branches of government basically just started investigating, coming after our company. And it was brutal.”
 
Any suggestion that jawboning by officials at the White House and in the agencies was purely a matter of advice should be laid to rest by the Judiciary Committee’s analysis and postings, as well as Zuckerberg’s description of being on the receiving end of this treatment. As Zuckerberg said, “you can’t censor that if it’s real legitimate information because it’s not ideologically convenient for you.”