Section 230 Explained: Platform Liability and Content Moderation

Section 230 is a provision of the Communications Decency Act that gives online platforms broad immunity from civil liability for content posted by their users, while also protecting their ability to moderate that content without becoming publishers. The statute, codified at 47 U.S.C. § 230, is the single sentence most often credited with making the modern consumer internet legally viable. Without it, every comment thread, product review, and social post would become a potential libel claim against the platform hosting it.

The law is also one of the most contested statutes in US technology policy. Reform bills from both parties have targeted it since at least 2018, the EU has built an entirely different regulatory model around its absence, and tech companies operating in both jurisdictions now run two parallel content-moderation regimes to stay compliant. The stakes of any change run from speech rights to startup viability, and the underlying statutory text is shorter than this paragraph.

The Text and Structure of Section 230

Section 230 of the Communications Decency Act establishes two core immunities: subsection (c)(1) shields platforms from publisher liability for user content, and subsection (c)(2) protects them from liability for good-faith content moderation decisions. The first clause, often called the twenty-six words that created the internet, states that no provider of an interactive computer service shall be treated as the publisher of information provided by another information content provider. The second clause, the Good Samaritan provision, extends immunity to platforms that voluntarily restrict access to material they consider obscene, harassing, or otherwise objectionable.

The Communications Decency Act was passed in 1996 as Title V of the Telecommunications Act. Most of the CDA was struck down by the Supreme Court in Reno v. ACLU (1997) on First Amendment grounds, but Section 230 survived and now operates as a freestanding immunity regime. The statutory text distinguishes between an interactive computer service, which hosts third-party material, and an information content provider, which originates it. A platform can be both, but only loses immunity for the specific content it originates.

The five terms below define how courts evaluate Section 230 claims, drawing on the statutory definitions in 47 U.S.C. § 230 at the Cornell Legal Information Institute.

Interactive computer service

Any information service, system, or access software provider that enables computer access by multiple users to a server. Covers social networks, forums, ISPs, web hosts, search engines, and cloud storage.

Information content provider

Any person or entity responsible, in whole or in part, for the creation or development of information. The originator of speech, not the host.

Publisher

Under traditional defamation law, an entity that exercises editorial control over content and is liable for what it publishes. Section 230(c)(1) blocks this classification for hosted user content.

Good Samaritan protection

The immunity in Section 230(c)(2) that allows a platform to remove objectionable user content in good faith without that act of moderation creating publisher liability.

Civil liability

Non-criminal legal exposure, including defamation, negligence, and consumer-protection claims. The statute bars this form of legal exposure for hosted user content; it does not bar federal criminal prosecution or intellectual property claims.

How Section 230 Shapes Content Moderation in Practice

Section 230's Good Samaritan clause allows platforms to remove content they find objectionable without triggering the publisher liability that would arise if they were deemed to have curated it. Before the law existed, US courts had reached the opposite conclusion on platform liability. In Stratton Oakmont v. Prodigy (1995), a state court ruled that a service moderating its forums had assumed editorial responsibility and was therefore liable for defamatory user posts it failed to remove. Congress passed the CDA the following year in part to overturn that outcome and remove the legal penalty for moderating.

The practical consequence is that platforms can write and enforce community guidelines without forfeiting immunity for everything else on the service. A platform that removes a hateful post does not become responsible for the millions of posts it leaves up. The clause is also the legal foundation for labeling, demonetization, age-gating, and shadow-banning, all of which involve moderation decisions that would otherwise look like editorial acts.

Two narrow carveouts limit the protection. FOSTA-SESTA, the 2018 amendment, removed immunity for content that knowingly facilitates sex trafficking. Federal criminal law has always sat outside the statute; the law only bars civil liability and state criminal claims that are inconsistent with it. Children's privacy enforcement under COPPA also runs around the immunity, since it targets the platform's own data collection practices rather than hosted user-generated content. The five scenarios below show where the line falls between protected moderation and statutory carveout, with reference to the original CDA text at Congress.gov.

1. Removing hate speech under community guidelines. Coverage applies. Good-faith removal of objectionable user-generated content is the canonical (c)(2) use case.
2. Labeling misinformation without removal. Coverage applies. Adding a label or interstitial is a moderation act that does not convert the platform into the information content provider for the underlying post.
3. Demonetizing a channel. Coverage applies. Restricting access to monetization tools is a service decision protected by (c)(2) and by the platform's First Amendment right to set terms.
4. Cooperating with law enforcement subpoenas. Immunity does not apply. Federal criminal law sits outside the statute, and a court order compelling production of user data is governed by the Stored Communications Act, not by the CDA.
5. Hosting content that facilitates sex trafficking. Immunity does not apply. The 2018 FOSTA-SESTA amendment carved an explicit exception for trafficking-related material, and the host can face both civil and criminal exposure.

Section 230 Reform Proposals and Alternatives

Section 230 has faced reform proposals from both parties, conservatives argue it enables viewpoint suppression without accountability, and progressives argue it shields platforms from responsibility for algorithmic amplification of harmful content. The result is a strange bipartisan pressure on a statute that almost no one wants to repeal outright but that almost everyone wants to narrow in scope. The platform liability debate has intensified as algorithmic amplification systems have grown more sophisticated and harder to separate analytically from neutral hosting. Two bills have anchored most of the recent reform debate: the SAFE TECH Act, sponsored by Senate Democrats, and the EARN IT Act, a bipartisan proposal focused on child sexual abuse material.

The SAFE TECH Act would strip immunity for paid content, including advertising, and would carve out claims involving stalking, harassment, civil rights violations, and wrongful death. The EARN IT Act conditions immunity on compliance with a federal commission's best practices for combating child sexual abuse material, and removes immunity for state-law claims involving such material. The FTC has separately framed platform accountability as an enforcement priority across consumer-protection and advertising rules, as described in its work on platform accountability. The table below compares the structural changes each proposal would make to the current statute.

The reform case rests on three operational arguments that recur across both bills.

1. Platforms face no consequence for algorithmically amplifying harmful content, since Section 230 immunity attaches to hosting rather than to recommendation systems.
2. Smaller companies cannot self-regulate against well-funded bad actors without a liability floor that distinguishes negligent hosting from active facilitation.
3. FOSTA-SESTA demonstrated that narrow statutory carveouts are workable without collapsing the broader immunity regime, even if the public-health evidence on its specific outcomes remains contested.

Section 230 vs the EU Digital Services Act

Section 230 and the EU Digital Services Act represent two divergent models of internet governance: the US statute assumes platform neutrality as the default, while the DSA imposes proactive due diligence obligations on very large online platforms. The DSA, Regulation (EU) 2022/2065, entered full application in February 2024, as confirmed by the official DSA text at EUR-Lex, and applies tiered duties based on platform size. Very large online platforms and very large online search engines, those with more than 45 million monthly EU users, face the strictest obligations: annual algorithmic audits, transparency reporting, mandatory risk assessments on systemic harms, and structured access for vetted researchers.

The contrast is structural. The law grants immunity by default and requires no proactive duty; the EU Digital Services Act imposes proactive duty by default and conditions safe-harbor on compliance with due diligence obligations. A platform serving both jurisdictions must run two policy regimes in parallel, which has implications for internet governance, hosting architecture, and where moderation review teams sit. The corresponding infrastructure decisions track the framework set out in data localization laws and their impact on cloud services. The five-row comparison below summarizes the headline differences, based on the official text of the DSA at EUR-Lex (linked above).

Implications of Repealing Section 230

Repealing Section 230 without a replacement framework would expose platforms to publisher-level platform liability for every piece of user-generated content, creating strong incentives for either hyper-moderation or exit from the US market. The legal calculation would shift overnight from policy judgment to litigation exposure modeling. Every hosted post would become a potential defamation, negligence, or consumer-protection claim, and the platform's only defense would run through the underlying merits of the user's speech rather than through statutory immunity.

The fallout would land unevenly across the market. Repeal scenarios intersect with adjacent questions of internet governance and open-internet policy, including the net neutrality policy debates over carrier-level neutrality, since both ultimately ask who bears responsibility for what flows across infrastructure. The five consequences below trace the most likely operational outcomes of a clean repeal.

1. Surge in preemptive content takedowns as platforms remove borderline material to limit jury exposure, shrinking the range of speech available online.
2. Smaller platforms unable to absorb litigation costs, consolidating the market toward incumbents with the legal budget to defend high volumes of user-generated content.
3. Chilling effect on political and minority-community speech, since the categories most likely to draw defamation or harassment suits also tend to be the most controversial.
4. International platforms restructuring US operations or geofencing US users, treating the country the way some now treat jurisdictions with hostile speech regimes.
5. US courts becoming the de facto global content moderation arbiters, since jury awards in US defamation cases would set the operational floor for what platforms tolerate worldwide.

Any serious reform debate has to answer two questions the current bills mostly skirt: what replaces the Good Samaritan clause, and how do smaller services survive the transition. Neither the SAFE TECH Act nor the EARN IT Act offers a clean answer, which is part of why neither has passed.

Frequently Asked Questions