Social media firms, as private actors, are not subject to the First Amendment. The extraordinary power of private companies to censor political speech, and the danger that government officials will enlist private firms to censor political opponents are serious social concerns.

In early July of 2021, former US President Donald Trump announced that he was filing class action lawsuits against the three largest social media companies—Facebook, Twitter, and YouTube­—as well as their Chief Executive Officers (CEOs), claiming that these companies had violated Trump’s, and his fellow plaintiffs’, rights to free speech protected by the First Amendment to the US Constitution. The lawsuits attack social media firms’ decisions to block, or to label as suspicious, a variety of posts by Trump and others, culminating in their decisions to ban Trump from their platforms entirely following the 6 January 2021 attack on the United States Capitol by a mob of Trump supporters, following a rally attended by President Trump.

The obvious problem with the Trump lawsuits is that the defendants in these actions, social media firms and their CEOs, are private actors not the government, but it is undisputed that the First Amendment only limits governmental restrictions on free speech. The lawsuits seek to get around this barrier in two ways. First, Trump claims that social media decisions to censor Trump and his followers were made under pressure from Democratic members of Congress, and in concert with government entities and officials such as the Center for Disease Control (CDC) and Dr Anthony Fauci. As such, the argument goes, the censorship is attributable to the government. Second, Trump claims that Section 230 of the Communications Decency Act, the federal statute authorising Internet platforms to restrict content on their platforms without fear of liability, itself violates the First Amendment by authorising censorship.

A word of explanation. Section 230—famously labeled “The Twenty-Six Words that Created the Internet”—was enacted by the US Congress in 1996, years before social media was even envisioned. It has two crucial substantive provisions. First, it states that internet platforms shall not be treated as publishers or speakers of content posted by third parties, thereby relieving platforms of legal liability for such content. Second, it states that platforms shall not be held legally liable for decisions made in good faith to remove or restrict content that, though constitutionally protected, is “objectionable” in various ways. This latter provision permits social media to block speech such as glorification of violence, pornography, and hate speech, all of which have been held to be protected by the First Amendment but are nonetheless obviously problematic.

There is almost no chance that these lawsuits will succeed. Insofar as they claim that social media firms worked in “concert” with government officials, the most that is actually alleged is the firms relied upon government sources such as the CDC and Dr Fauci in determining what kinds of factual claims to label as COVID-19 or vaccine misinformation, and offered their services to spread the government’s health messaging. It is inconceivable that such public-private coordination, which is extremely common in a wide range of areas, can transform a private party such as Facebook into a state actor. The claim that members of Congress coerced social media firms into censorship is even weaker. First, there is simply no evidence that the defendants, some of the wealthiest corporations and individuals in the world, were actually coerced into censorship (though some politicians clearly have been urging social media to restrict more content). Further, if they were coerced, that would make social media firms the victims of government coercion rather than the wrongdoers, making Trump’s decision to sue the victims decidedly odd.

Nor, contrary to the Trump lawsuits, is Section 230 itself unconstitutional. The second part of Section 230 authorising platforms to moderate content in good faith without fear of liability, which is the focus of the Trump lawsuits, reinforces First Amendment values. This is because social media firms possess First Amendment rights to control what content appears on their platforms, just as newspapers have well-established First Amendment rights to publish, or refuse to publish, whatever they want free of governmental interference. Admittedly, social media editorial rights are probably somewhat weaker those of newspapers, for a variety of reasons. Nonetheless, platforms possess a prima facie right to control the content they host, subject only to narrow regulations.

Putting this all together, it follows that far from vindicating the First Amendment, as the Trump lawsuits claim they are doing, what Trump is actually seeking to do via litigation is to eviscerate social media firms’ First Amendment rights to control the content on their platforms.  That is why the lawsuits not only will fail, but should fail.

Although the Trump lawsuits are legally weak, they do highlight some serious concerns. First and foremost, it is extraordinary that three private firms were able to effectively silence a sitting president of the United States (albeit an extraordinary President) by deplatforming him. Such private power over public discourse should make many people, including foreign leaders, very nervous. Throughout the world today, social media is the key vehicle for political debate. That three individuals, all sitting in the San Francisco Bay Area (where I currently sit), should have the power to silence almost any world leader is staggering and troubling.

But government regulation of platforms, even if constitutional (which I doubt), is not the solution. For one thing, such regulatory authority will almost certainly be exercised in biased ways—it is surely no coincidence that the loudest advocates for controlling social media are often quasi-authoritarian leaders such as Trump. In addition, reducing social media firms’ ability to restrict content, as the Trump lawsuits seek to do, is a supremely terrible idea. It would mean a lot more violence, a lot more pornography, and a lot more hate speech on social media, at least in the United States where all such speech is constitutionally protected. There are sound reasons why Facebook and Twitter want to keep “lawful but awful” content off their platforms, because if such content proliferated all but the worst users would flee. Such an outcome would be to almost no one’s benefit. As a result, at present our best option is almost certainly to leave control over content in the hands of the platforms, subject to public oversight and criticism.

The other, lurking danger is that government officials will pressure or coerce social media firms to silence their critics and opponents. Especially in countries veering towards authoritarianism, including democratic India, this risk is very real and indeed has already manifested itself. Contrary to the Trump lawsuits, such behavior has not yet been evident in the United States (when President Biden and Facebook entered into a public dispute over vaccine disinformation, notably it was President Biden who backed down) or other functional democracies, but the temptation to suborn social media is obvious. Again, however, government regulation is obviously not the correct solution here, since such authority would merely give new levers for officials to impose pressure. Again, therefore, transparency and an engaged citizenry seem for now the best we can hope for.

Ashutosh Bhagwat is a Distinguished Professor of Law at the University of California, Davis School of Law. He holds a B.A., summa cum laude, from Yale University and a J.D. with Honors from the University of Chicago. Professor Bhagwat has written extensively about the First Amendment to the U.S. Constitution, with a particular focus on the relationship between First Amendment rights and democracy, as well as the impact of new technologies on democratic discourse. His most recent book is Our Democratic First Amendment.

This article is published under a Creative Commons License and may be republished with attribution.