For nearly three decades, Section 230 of the Communications Decency Act has provided internet platforms with broad immunity from liability for content posted by their users. But a new wave of litigation is testing the limits of that protection, raising a fundamental question: When social media companies use algorithms to recommend, amplify, and prioritize content, are they engaging in protected speech or potentially actionable conduct?
The answer could reshape the legal landscape for social media companies, online publishers, and anyone who relies on algorithms to distribute information.
Section 230 generally protects online platforms from being treated as the publishers or speakers of information provided by third parties. Courts have traditionally interpreted that protection broadly, recognizing that platforms must make decisions about how to organize, filter, and display user-generated content. But today’s social media platforms do much more than simply host information. Their algorithms determine which posts users see, recommend videos based on individual preferences, and amplify content designed to maximize engagement. Plaintiffs increasingly argue that these activities constitute affirmative conduct by the platforms themselves, rather than merely the publication of third-party speech. That distinction could prove critical.
In Gonzalez v. Google LLC (2023), the Supreme Court had an opportunity to decide whether Section 230 protects algorithmic recommendations. The case involved allegations that YouTube’s recommendation algorithms promoted terrorist content. Rather than resolve the issue, the Court declined to address Section 230’s scope, leaving the central question unanswered. The Court subsequently addressed a related constitutional issue in Moody v. NetChoice, LLC (2024), recognizing that platforms’ decisions about selecting, organizing, and presenting third-party content can constitute protected editorial expression under the First Amendment. Together, these decisions establish an important distinction: even when Section 230 immunity is uncertain, algorithmic content curation may receive independent constitutional protection.
State legislatures have increasingly attempted to regulate social media platforms through restrictions on addictive design features, protections for minors, and requirements governing content recommendations. California and New York have been at the forefront of these efforts. For example, New York’s “SAFE For Kids Act” attempts to regulate platforms that provide an “addictive feed” to users under eighteen years old. Technology companies have responded with constitutional challenges, arguing that such laws interfere with their editorial discretion and compel speech. The legal question is whether these statutes regulate harmful product design or impermissibly burden protected expression. That distinction matters because not every algorithmic function necessarily constitutes speech. A recommendation system that selects and ranks expressive content may implicate the First Amendment, while features involving notifications, data collection, or user-interface design may present different constitutional questions.
The next generation of Section 230 litigation will likely turn on how courts characterize the conduct at issue.
If algorithmic recommendations are treated as traditional publishing decisions, platforms may enjoy both statutory immunity and First Amendment protection. If courts instead characterize particular algorithmic features as independently harmful conduct, plaintiffs may have a path around Section 230, although constitutional defenses could remain. For media companies, the stakes extend well beyond social media. News aggregators, streaming services, search engines, and digital publishers increasingly rely on algorithms to select and distribute content. The central question is no longer simply who created the content. It is whether the technology used to distribute that content is itself an exercise of editorial judgment.
The answer may determine the future boundaries of online free expression.