Online content moderation is an exercise of free speech

The conduct-vs.-speech game is lose-lose, especially when applied to the internet.

The Supreme Court will rule this year on the constitutionality of a pair of bills from Florida and Texas that would drastically reduce social media platforms’ discretion in removing user content.

The First Amendment prohibits government interference with the editorial discretion of private entities. To circumvent this First Amendment protection online, those opposed to Big Tech have coalesced around a legal strategy: labeling speech as conduct. And if states like Florida manage to convince the Supreme Court of this distinction, everyone will lose — because two can play that game.

Florida’s brief to the Supreme Court in Moody v. NetChoice is the latest attempt to superimpose the conduct-vs.-speech construct onto social media. The state argues that its law, S.B. 7072, “regulate[s] conduct, not speech or expression.” Specifically, Florida claims to regulate the “conduct” of “hosting … third-party speech.”

Florida’s argument relies on Rumsfeld v. FAIR, a 2006 Supreme Court case upholding a federal statute that requires law schools to host military recruiters in order to receive government funding, which became a flashpoint during the era of “don’t ask, don’t tell.” The court reasoned that, “A law school’s decision to allow recruiters on campus is not inherently expressive.” The statute regulated “conduct, not speech.” It affected “what law schools must do — afford equal access to military recruiters — not what they may or may not say.”

Citing FAIR, Florida asserts: “In hosting billions of speakers and petabytes of content, the platforms are engaged in business activity — conduct — that may be regulated in the public interest. The First Amendment does not afford those who host third-party speech a right to silence the hosted speakers or to treat them arbitrarily.”

But Florida plays fast and loose with the distinction between what is expressive and what is not. The state accuses social media platforms of “silencing or mistreating a disfavored few among that cacophony of voices” they host. Florida concedes, however, that this so-called “hosting” includes “the presentation of content on the platforms to users.” The state dismisses all this activity as “nonexpressive conduct.”

Any reasonable person would recognize the activity of taking content down, leaving content up and showing curated content to users as content moderation. And content moderation is inherently expressive, conveying the policies and values of the social media companies. Consequently, content moderation more closely resembles speech than conduct. The plaintiff and a host of amici, including First Amendment experts, agree.

It’s worth considering what it will look like if the Supreme Court agrees with Florida that content moderation practices constitute regulable conduct under the First Amendment. Activists on both the right and left will clamor to weaponize the decision to advance their culture war goals, selectively applying Rumsfeld v. FAIR to punish content moderation decisions that negatively affect speech they support.

It’s already happening. Florida and Texas passed legislation aimed at preventing social media platforms from removing content posted by conservatives, resulting in the NetChoice litigation.

Democrats are playing the game too. In 303 Creative v. Elenis, the Supreme Court held that Colorado cannot compel a website designer to create expressive designs speaking messages “inconsistent with her belief that marriage should be reserved to unions between one man and one woman.” While the case was before the court, the Biden administration filed an amicus brief in support of Colorado, arguing that the First Amendment “does not entitle a commercial wedding-website service to categorically refuse to serve same-sex couples.”

Invoking FAIR, the Biden administration argued that the Colorado law in question regulates “conduct” while “impos[ing] only an incidental burden” on expression and speech. The Biden brief cited FAIR more than two dozen times.

FAIR was authored by the conservative Chief Justice John Roberts, but he seemed to downplay the breadth of the decision at the oral argument in 303 Creative.There, Roberts stated that FAIR involved “providing rooms,” holding merely that “empty rooms don’t speak.”

The conduct-vs.-speech game is lose-lose, especially when applied to the internet. If the Supreme Court sides with Florida, nobody will get what they want. When social media platforms remove posts containing sexually graphic images of Hunter Biden, Republicans will call it conduct and demand the nudes stay up. Similarly, when social media services remove or block pro-Hamas posts, Palestinian activists will call it conduct and demand the platforms host their content. FAIR will always work both ways.

Eric Goldman, a leading expert on the First Amendment and online speech, aptly summarized the game at play: “If [the Florida and Texas] laws survive the various legal challenges, we can anticipate a roughly infinite number of lawsuits over content moderation decisions, each of which necessarily prioritized some content over others.”

Clear-eyed Americans should reject the notion that the government can or should regulate speech on social media by labeling it conduct. We must not cheerlead lawmakers as they game out strategies for curtailing speech rights over senselessly politicized content moderation decisions.

Andy Jung is associate counsel at TechFreedom, a nonprofit, nonpartisan think tank focused on technology law and policy.