Skip to main content

How America turned against the First Amendment

Moderation laws. Book bans. Courts that keep getting played. America’s politicians are tired of the First Amendment getting in their way, and no one seems to care.

Share this story

Brightly colored large emojis and general internet iconography crowd a neoclassical building with a chat bubble icon at the top.
Illustration by Shira Inbar
Illustration by Shira Inbar

The First Amendment is one of America’s most distinctive pieces of law. But as the midterm — and, quickly, the presidential — elections approach, free speech has become one of the most fundamentally imperiled parts of American life. Both sides of the political aisle are attacking long-accepted principles of speech law, frequently in ways that are both logically incoherent and deeply concerning.

Thanks to the internet, the free speech clause of the First Amendment is bearing a level of strain that few other legal doctrines face in modern times. The modern internet touches virtually every aspect of our lives, and at some level, virtually every part of the internet is speech. Even more than many other large industries, its lack of geographical boundaries has eroded the barriers between local governments: a law regulating websites in Florida might dictate how businesses behave in California or Europe, while conversely, a web business that runs afoul of one country’s laws can quickly hide abroad. And a human-run legal system struggles to keep up with the vast scale of automated digital connections.

The concerns and concern-trolling around “cancel culture” and “illiberalism” have never been louder. Yet, at the same time, the freedom of speech — the actual civil liberty protected by the First Amendment, not just a theoretical Enlightenment value — is being gutted by legislatures and judges.

Rather than seriously grappling with technology’s effects on democracy, many lawmakers and courts have channeled a cultural backlash against “Big Tech” into a series of glib sound bites and political warfare. Scratch the surface of supposedly “bipartisan” internet regulation, and you’ll find a mess of mutually exclusive demands fueled by reflexive outrage. Some of the people most vocally defending the First Amendment are the ones most open to dismantling it — without even admitting that they’re doing so.


Section 230 is a smoke screen

Virtually every American politician professes to love the First Amendment. Many of them profess to hate another law: Section 230 of the Communications Decency Act. But the more they say about 230, the clearer it becomes that they actually hate the First Amendment and think Section 230 is just fine. 

The heart of Section 230 is famously just 26 words:

No provider or user of an interactive computer service shall be treated as the publisher or speaker of any information provided by another information content provider.

The law was passed in 1996, and courts have interpreted it expansively since then. It effectively means that web services — as well as newspapers, gossip blogs, listserv operators, and other parties — can’t be sued for hosting or reposting somebody else’s illegal speech. The law was passed after a pair of seemingly contradictory defamation cases, but it’s been found to cover everything from harassment to gun sales. In addition, it means courts can dismiss most lawsuits over web platform moderation, particularly since there’s a second clause protecting the removal of “objectionable” content.

The oft-neglected key here is illegal speech. Many well-deserved critiques of the internet and social media — that it helps spread false stories about covid or QAnon-style Satanic panics, that it lets huge crowds of people dogpile teachers or nurses with angry messages, or that it facilitates hate speech at a large scale — don’t actually involve illegal speech. There are a few cases still in flux, like the defamation lawsuits against Fox News for provably false and unsupported statements about voting machine manufacturers. But defamation is a difficult bar to meet. Joe Biden, for instance, claimed on the campaign trail that Section 230 let Facebook host disinformation. Last year, he took Facebook to task for “killing people” by allowing the spread of covid vaccine misinformation; shortly thereafter, Democratic senator Amy Klobuchar proposed stripping Section 230 protections for health misinformation.

Removing Section 230 protections is a sneaky way for politicians to get around the First Amendment

But making false claims about pandemic science isn’t necessarily illegal, so repealing Section 230 wouldn’t suddenly make companies remove misinformation. There’s a good reason why the First Amendment protects shaky scientific claims. Think of how constantly our early understanding of covid shifted — and now imagine researchers and news outlets getting sued for publishing good-faith assumptions that were later proven incorrect, like covid not being airborne.

Removing Section 230 protections is a sneaky way for politicians to get around the First Amendment. Without 230, the cost of operating a social media site in the United States would skyrocket due to litigation. Unable to invoke a straightforward 230 defense, sites could face protracted lawsuits over even unambiguously legal content. And when it comes to categories of speech that are dicier, web platforms would be incentivized to remove posts that might be illegal — anything from unfavorable restaurant reviews to MeToo allegations — even if they would have ultimately prevailed in court. All of this would burn time and money in perhaps existential ways. It’s no wonder platform operators do what it takes to keep 230 alive. When politicians gripe, the platforms respond.

Threatening to repeal 230 is a shakedown racket, a way for lawmakers to quietly put their thumb on the scale — a back door to imposing government speech regulations.  

There are some proposed changes to speech law that are upfront about their aims, like New York Attorney General Letitia James’ call to ban distributing live videos filmed by mass shooters. Legal experts like Danielle Citron have also proposed fixing specific problems created by Section 230, like its de facto protections for small sites that solicit nonconsensual pornography or other illegal content. There are serious criticisms of these approaches, but they’re honest attempts to address real legal tradeoffs.

When lawmakers do take up serious issues around platform regulation, it’s often with the blatant ulterior motive of punishing “Big Tech” for perceived political misdeeds. Debate over the EARN IT Act targeting child sexual abuse material, for instance, was peppered with claims that Section 230 was a gift to Facebook or Google — when groups like NCMEC acknowledged that the popular “Big Tech” targets were some of the most responsive at reporting CSAM. Omnibus Section 230 reform bills are bizarre chimeras from lawmakers with conflicting goals, pushing sites to moderate more and less at the same time.

The thing is, these complaints get a big thing right: in an era of unprecedented mass communication, it’s easier than ever to hurt people with illegal and legal speech. But the issue is far bigger and more complicated than encouraging more people to sue Facebook — because, in fact, the legal system has become part of the problem.


The legal system wasn’t built for bad faith at scale

Last month, longtime fabulist Alex Jones was hit with a judgment for nearly $1 billion for lying about the 2012 shooting at Sandy Hook, Connecticut. His behavior — and I say this as a reporter, one of the groups most leery of libel lawsuits — is exactly why defamation isn’t protected by the First Amendment. Jones targeted private citizens who’d suffered a horrific loss with absurd, protracted, totally unsupported claims. His lies caused tangible harm that continues a decade later, and in theory, the damages outweigh even the tens of millions of dollars he made off those claims.

It’s also not clear whether it matters. Jones declared corporate bankruptcy during the procedure, tying up much of his money indefinitely and leaving Sandy Hook families struggling to chase it. He treated the court proceedings contemptuously and used them to hawk dubious health supplements to his followers. Legal fees and damages have almost certainly hurt his finances, but the legal system has conspicuously failed to meaningfully change his behavior. If anything, it provided yet another platform for him to declare himself a martyr.

Contrast this with the year’s other big defamation case: Johnny Depp’s lawsuit against Amber Heard, who had identified publicly as a victim of abuse (implicitly at the hands of Depp). Amber Heard’s case was less cut-and-dried than Jones’, but she lacked Jones’ shamelessness or social media acumen. The case turned into a ritual public humiliation of Heard — fueled partly by the incentives of social media but also by courts’ utter failure to respond to the way that things like livestreams contributed to the media circus. Defamation claims can meaningfully hurt people who have to maintain a reputation, while the worst offenders are already beyond shame.

The First Amendment doesn’t work if the legal system doesn’t work

Everything is content now, and the legal system is no exception. Influencers — up to the former president — have honed the art of filing absurd speech-related legal complaints to grab headlines from an ecosystem of media reporters. They’ve figured out how to barrage targets with lawsuits that are laughably weak. But they don’t even want to win. They want to gin up social media outrage, incite harassment, and discourage people who can’t pay legal fees from speaking out. There’s a welcome and helpful movement to address bad-faith lawsuits through anti-SLAPP (or Strategic Lawsuits Against Public Participation) laws, but even these only partially address what makes the state of US speech so poisoned.

To put it bluntly, the First Amendment doesn’t work if the legal system doesn’t work. Arguing over the rare exceptions to free speech doesn’t matter if people can’t be meaningfully censured for serious violations or if verdicts are vestigial afterthoughts in cases filed mostly for clout. And it’s especially useless if the courts themselves won’t take it seriously.

Emojis and internet iconography placed in the scales of justice.
Illustration by Shira Inbar
Video

I guess corporations aren’t people again

Up until this point, I’ve almost exclusively addressed Democratic and bipartisan proposals to reform Section 230 because those at least have some shred of substance to them. 

Republican-proposed speech reforms are ludicrously, bizarrely bad. We’ve learned just how bad over the past year, after Republican legislatures in Texas and Florida passed bills effectively banning social media moderation because Facebook and Twitter were using it to ban some posts from conservative politicians, among countless other pieces of content.

As it stands, the First Amendment should almost certainly render these bans unconstitutional. They are government speech regulations! But while an appeals court blocked Florida’s law, Texas’ Fifth Circuit Court of Appeals threw a wrench in the works with a bizarre surprise decision to uphold its law without explaining its reasoning. Months later, that court actually published its opinion, which legal commentator Ken White called “the most angrily incoherent First Amendment decision I think I’ve ever read.”

Republican-proposed speech reforms are ludicrously, bizarrely bad

The Supreme Court temporarily blocked the Texas law, but its recent statements on speech haven’t been terribly reassuring. It’s almost certain to take up either the Texas or Florida case, and the case will likely be heard by a court that includes Clarence Thomas, who’s gone out of his way to argue that the government should be able to treat Twitter like a public utility. (Leave aside that conservatives previously raged against the idea of treating ISPs like a public utility in order to regulate them; it will make your brain hurt.)

Thomas, as well as two other conservative justices, voted against putting the law on hold. (Liberal Justice Elena Kagan did, too, but some have interpreted her vote as a protest against the “shadow docket” where the ruling happened.)

It’s deeply worrying that the Supreme Court seems so open to gutting a long-standing First Amendment principle: most people and companies shouldn’t have to publish speech they find abhorrent. 

Look, I’ll fully admit this: it’s weird to be in the position of arguing against limits on corporate power to control speech. Facebook, TikTok, Twitter, and other companies all play a huge role in public discourse and exercise a huge amount of influence over how Americans can connect with each other. It’s getting harder and harder to talk to other people in a way that’s not monitored and approved by an increasingly small number of companies.

The rules are transparently rigged to punish political targets at the expense of basic consistency

But only a useful idiot would support the laws in Texas and Florida on those grounds. The rules are transparently rigged to punish political targets at the expense of basic consistency. They attack “Big Tech” platforms for their power, conveniently ignoring the near-monopolies of other companies like internet service providers, who control the chokepoints letting anyone access those platforms. There is no saving a movement so intellectually bankrupt that it exempted media juggernaut Disney from speech laws because of its spending power in Florida, then subsequently proposed blowing up the entire copyright system to punish the company for stepping out of line.

And even as they rant about tech platform censorship, many of the same politicians are trying to effectively ban children from finding media that acknowledges the existence of trans, gay, or gender-nonconforming people. On top of getting books pulled from schools and libraries, Republican state delegate in Virginia dug up a rarely used obscenity law to stop Barnes & Noble from selling the graphic memoir Gender Queer and the young adult novel A Court of Mist and Fury — a suit that, in a victory for a functional American court system, was thrown out earlier this year. A disingenuous panic over “grooming” doesn’t only affect LGBTQ Americans. Even as Texas is trying to stop Facebook from kicking off violent insurrectionists, it’s suing Netflix for distributing the Sundance-screened film Cuties under a constitutionally dubious law against “child erotica.”

Republican lawmakers are even attacking spam filters as biased

These attacks on the First Amendment are already affecting some of the most vulnerable Americans, but they have far-reaching implications for everyone. The Texas and Florida laws rules aren’t even written carefully enough to ensure they only apply to “Big Tech.” Under some interpretations, Texas’ law means Wikipedia wouldn’t be allowed to remove edits that violate its standards. Republican lawmakers are even attacking spam filters as biased, so the effects aren’t just theoretical — if courts rule the wrong way, your inbox may be about to get a lot messier.

But once again, there’s a real and meaningful tradeoff here: if you take the First Amendment at its broadest possible reading, virtually all software code is speech, leaving software-based services impossible to regulate. Airbnb and Amazon have both used Section 230 to defend against claims of providing faulty physical goods and services, an approach that hasn’t always worked but that remains open for companies whose core services have little to do with speech, just software. 

Tech freedom advocates have fought for years against laws that would stifle online communication, a project based on the assumption that this communication is a social good. The limits of this assumption have never been clearer, and the backlash threatens to make things even worse.


None of us are as cruel as all of us

The Awl co-founder Alex Balk once wrote: “Everything you hate about the Internet is actually everything you hate about people.” 

Balk’s Law is obviously an oversimplification. Internet platforms change us — they incentivize specific kinds of posts, subjects, linguistic quirks, and interpersonal dynamics. But still, the internet is humanity at scale, crammed into spaces owned by a few powerful companies. And it turns out humanity at scale can be unbelievably ugly. Vicious abuse might come from one person, or it might be spread out into a campaign of threats, lies, or stochastic terrorism involving thousands of different people, none of it quite rising to the level of a viable legal case. 

The internet is humanity at scale, crammed into spaces owned by a few powerful companies

Platform moderation has become one of the few enforcement mechanisms to punish bad behavior. When bad behavior goes unpunished, unaddressed, and spirals out into something worse and worse, people look for something to blame. And the reality is that no politician can blame the First Amendment, so they blame 230 instead. That’s the lesson, the takeaway: whenever politicians talk about regulating Big Tech or changing 230, they are almost always talking about imperiling the First Amendment.

The bigger issue, though not the only one, is that the internet allows people to speak to each other at a scale unprecedented in human history. The shortcomings and tradeoffs of the laws governing that speech have never been so evident, and their troublesome edge cases never so numerous. And instead of trying to reckon with a new world, the people who make and enforce those laws have abdicated their principles and responsibilities in favor of wielding raw power — and, often, abdicating a lot of their common sense as well.

The First Amendment is over 200 years old. It might remain the law on paper for a long time to come, but for the average American, the era of government speech regulations is already here, and things are about to get darker than ever.

Correction: Cuties was screened at the Sundance Film Festival, not Cannes as previously stated, and the Texas ruling came from the Fifth Circuit Court of Appeals, not the Fourth. We regret the error.