Does Free Speech Apply to Social Media?

The straightforward legal answer is that the First Amendment prohibits government censorship of speech — including on social media — but does not require private social media platforms to host any particular speech. When the government tells Twitter to remove a user's account, or passes a law requiring Facebook to carry certain content, the First Amendment is implicated. When Facebook or YouTube itself decides to remove content or ban users, it is not the government acting, and the First Amendment does not apply.

The Supreme Court confirmed this framework in Moody v. NetChoice (2024), which addressed Florida and Texas laws attempting to prohibit social media platforms from removing or restricting certain content. Both states argued that dominant platforms had become so powerful that they functioned like utilities or public forums, requiring government-imposed neutrality. The Court rejected this reasoning, reaffirming that private platforms have their own First Amendment rights to make editorial choices about what content they host and how they organize it. Platforms are more like newspapers or bookstores — editorial curators — than telephone companies or public utilities.

The practical implication is that users have no First Amendment right to a Twitter account, a Facebook page, or a YouTube channel. Platforms can suspend or terminate accounts based on their content policies, community standards, or even arbitrary decisions, without triggering constitutional liability. Users' recourse against platform censorship is not legal but market-based: leaving for competing platforms, advocating for better policies, or supporting regulatory reforms through the political process.

The foundational case establishing the internet as a maximally protected speech environment is Reno v. American Civil Liberties Union (1997), in which the Supreme Court struck down the Communications Decency Act's indecency provisions as applied to the internet. The case established that the internet does not receive the reduced First Amendment protection that applies to broadcast television and radio (the Pacifica doctrine), but rather the full First Amendment protection afforded to print media. Justice Stevens's majority opinion described the internet as 'a vast democratic forum' and rejected the analogy to broadcasting that would have supported content regulation.

Reno set the stage for the internet's development as a relatively unregulated speech environment. Section 230 of the Communications Decency Act, enacted the year before Reno was decided, provided the practical foundation: by immunizing platforms from liability for user-generated content and from liability for good-faith content moderation decisions, Section 230 enabled platforms to host enormous quantities of user speech without either becoming publishers in the traditional legal sense or being crushed by liability for the speech they hosted.

The combination of Reno's First Amendment framework and Section 230's immunity created the conditions for the development of social media as we know it — a largely self-regulated speech environment where platforms set their own content policies without government dictating what they must remove or retain. The question of whether this model remains appropriate as a handful of platforms have come to dominate public discourse — and whether the Reno framework fits an internet where power is concentrated in a few dominant companies rather than dispersed across countless websites — is at the center of current debates about social media regulation.

Section 230 of the Communications Decency Act provides the legal foundation for how social media platforms operate. It has two key provisions: (1) platforms are not treated as publishers or speakers for user-generated content, meaning they cannot be held liable for what their users post; and (2) platforms are not liable for good-faith efforts to restrict 'objectionable' content, protecting them from claims that moderation decisions make them responsible for content they retain. Together, these provisions allow platforms to host enormous amounts of user content while moderating some of it, without the legal exposure that would apply to a traditional publisher that selects and edits what it publishes.

Section 230's critics argue that it has enabled platforms to profit from harmful content — harassment, disinformation, radicalization — while immunizing them from the liability that would incentivize better behavior. The lack of liability, in this view, means platforms have insufficient incentive to invest in content moderation, and the combination of network effects and immunity has allowed dominant platforms to grow without accountability. Proposed Section 230 reforms range from narrowing the immunity for certain categories of content (CSAM, terrorism, health misinformation) to eliminating the immunity entirely for large platforms.

Section 230's defenders argue that any significant narrowing of the immunity would have severe consequences for online speech. Without liability protection, platforms would face a choice between vastly over-moderating (removing anything that might generate a lawsuit) or hosting no user content at all. The most likely result would be massive consolidation — only the largest platforms with the most resources to manage liability risk could survive — combined with aggressive over-moderation to minimize legal exposure. Smaller platforms, where much of the internet's diverse speech occurs, would be most at risk.

Moody v. NetChoice (2024) represented the Supreme Court's most significant engagement with social media and the First Amendment in recent years. The cases arose from Florida's SB 7072 and Texas's HB 20, both passed in 2021 after prominent conservatives were banned or restricted on major platforms following the January 6 Capitol riot. Both laws attempted to prohibit platforms from removing or restricting content based on political viewpoint, arguing that dominant platforms had become essential public infrastructure that could not discriminate based on viewpoint.

The Supreme Court vacated the lower court decisions and remanded, but in doing so articulated a framework that strongly protects platform editorial discretion. Justice Kagan's opinion for a partial majority held that platforms engage in editorial judgment when they compile, curate, and arrange content — activities that are protected by the First Amendment. When a platform decides what content to feature, what to remove, and how to organize its feeds, it is making expressive choices that the government cannot dictate without implicating the First Amendment.

The Moody decision has significant implications for future social media regulation. It places limits on how far governments can go in compelling platforms to carry speech they would prefer not to host. It also suggests that platform content moderation decisions — even when they affect millions of users — are exercises of editorial discretion rather than public utility-style services that government can mandate. The decision did not resolve every question about social media regulation, and future cases will define the boundaries between permitted platform editorial choices and government interests in ensuring open public discourse.

Even if social media platforms have no First Amendment obligations as a matter of constitutional law, the concentration of public discourse on a small number of dominant platforms raises fundamental democratic questions. When the vast majority of political discussion, news dissemination, and civil society organizing occurs on platforms controlled by a handful of private companies, those companies' content moderation decisions effectively determine the shape of public discourse. A Twitter ban affects a public figure's ability to communicate with their supporters in ways that a bookstore's refusal to stock a book does not, because Twitter has a de facto monopoly on a certain kind of real-time political communication.

The 'public forum' doctrine — which requires government to allow expression in traditional and designated public forums regardless of viewpoint — does not apply to private platforms under current First Amendment law. But the underlying rationale — that democratic participation requires spaces where all viewpoints can compete — arguably applies to the de facto public forums that dominant platforms have become. Whether the First Amendment's public forum doctrine should be extended to private platforms with sufficient monopoly power, or whether equivalent protections should be created through statute, is an active debate in First Amendment scholarship.

International approaches offer different models. The EU's Digital Services Act imposes obligations on 'very large' platforms to provide systemic risk assessments, researcher data access, and transparency about moderation decisions — without requiring neutral carriage of all content. The UK's Online Safety Act creates duty-of-care obligations for platforms regarding harmful content. These approaches attempt to impose accountability on platform power without either requiring neutral carriage of all speech or eliminating platforms' editorial discretion entirely.

The future of social media's relationship with free speech will be shaped significantly by AI — both AI as a tool for content moderation and AI as a source of content that stresses existing regulatory frameworks. AI moderation systems make billions of decisions daily about what content to allow, restrict, or amplify — decisions that aggregate into a comprehensive shaping of what speech users see and can say. As these systems become more sophisticated, the gap between nominally allowed speech and algorithmically amplified speech will grow: content can be permitted but effectively suppressed through reduced distribution, while content can be allowed and effectively promoted through algorithmic preference.

AI-generated content — bots, synthetic personas, AI-written posts and comments — is straining platforms' abilities to maintain authentic human discourse. Platform terms of service generally prohibit coordinated inauthentic behavior, but AI-generated content from individual users is harder to define and regulate. When a user employs an AI assistant to help draft social media posts, is that authentic human speech or AI-generated content subject to different standards? The line between AI-assisted and AI-generated expression will be a major regulatory question for social media platforms.

Government regulation of AI in social media contexts raises its own First Amendment concerns. Requirements that platforms label AI-generated content, mandates that platforms reduce algorithmic amplification of certain content categories, and obligations to provide algorithmic transparency all interact with platforms' First Amendment editorial rights in ways that courts have not yet fully addressed. The Moody framework — protecting platform editorial discretion — may limit what government can require of platforms even when the goal is ensuring authentic public discourse rather than promoting any particular viewpoint.