What Is Section 230, and Why Does Everyone Fight About It

Section 230 of the Communications Act is frequently called the most important law for the internet, and also one of the most misunderstood. Passed in 1996, its core is just 26 words. Yet those words sit at the center of a fight that has now run for a decade, with new repeal bills arriving in Congress as recently as September 2026. Supporters call it the foundation of online speech. Critics call it a shield for harmful content. Both sides are describing the same statute, so it is worth reading what it actually says before joining the argument. This article is explanatory only, not legal advice.

What the law actually says

Section 230 (47 U.S.C. 230) has two provisions that do most of the work. The first, subsection (c)(1), states that no provider or user of an interactive computer service shall be treated as the publisher or speaker of information provided by someone else. In plain terms: if a user posts a defamatory review on a forum, the forum generally cannot be sued as though it had written the review itself. The person who posted it can still be sued. The platform is treated as the venue, not the author.

The second, subsection (c)(2), is the "Good Samaritan" clause. It protects platforms from liability for good-faith efforts to restrict access to material they consider obscene, violent, harassing, or otherwise objectionable. This matters because it resolves a perverse incentive created by two early court cases. In Cubby v. CompuServe (1991), a service that did not moderate was treated like a bookstore and avoided liability. In Stratton Oakmont v. Prodigy (1995), a service that did moderate was treated like a publisher and held liable. The lesson platforms learned was: never moderate, or you become responsible for everything. Section 230 reversed that lesson, so that cleaning up a forum does not make the forum legally liable for what remains.

The immunity has defined exceptions. It does not cover federal criminal liability, and it does not cover intellectual property claims, which is why copyright takedowns run through a separate system. Congress narrowed it further in 2018 with FOSTA-SESTA, which removed the shield for knowing facilitation of sex trafficking.

Section 230 does not protect what you post. It protects the platform that hosts it, and that distinction is where most of the debate goes wrong.

What it does not say

Much of the public argument is about things the statute never contained. It does not protect the original speaker: if you post something illegal, Section 230 is no defense for you. It does not require political neutrality or balanced viewpoints; there is no such condition in the text, whatever various proposals have suggested adding. It does not prevent platforms from moderating; on the contrary, subsection (c)(2) was written to encourage moderation by removing the legal risk of doing it.

It also does not say platforms can never be liable for anything. Courts have generally read the immunity broadly, which is itself part of the controversy: critics argue judges stretched a 1996 law, written for dial-up bulletin boards, to cover modern algorithmic feeds that actively amplify content. Defenders respond that the text draws no distinction between chronological and algorithmic presentation, and that the venue-versus-author logic holds either way.

Why everyone fights about it

The repeal camp argues, in essence, that the bargain has broken down. Platforms now profit enormously from engagement-driven feeds, and immunity removes the financial incentive to deal with harassment, scams, nonconsensual imagery, drug sales, and misinformation spreading on their services. From this view, a law meant to protect fledgling forums in 1996 now subsidizes the business models of the world's largest companies, and the exceptions carved out so far have not kept up with algorithmic amplification.

The keep-it camp argues that repeal would not produce a nicer internet, but a quieter one. Without immunity, every user post becomes a potential lawsuit, and platforms would respond with aggressive filtering, or by shutting down user content entirely. Small forums, open-source projects, and startups could not afford the legal departments that giants can, so the likely winners of repeal would be the biggest incumbents, not their critics. User speech itself is at stake in this telling: ordinary people can only speak at scale because platforms are willing to host them.

Both arguments contain a real mechanism, which is why the fight never settles. It is a genuine tradeoff between accountability for harms and the conditions that make open platforms possible, and where you land depends on which risk you fear more.

The reform proposals as of October 2026

Congress has introduced several Section 230 bills in the current session, and the most striking are the straight-repeal efforts. The Sunset Section 230 Act began in the Senate as S.3546, introduced on 17 December 2025 by Senator Graham with a bipartisan group of cosponsors including Senators Durbin, Grassley, Whitehouse, Hawley, Klobuchar, Blackburn, Blumenthal, Moody, and Welch, and referred to the Committee on Commerce, Science, and Transportation. A House version, H.R. 10332, was introduced on 10 September 2026 by Representatives DeSaulnier and Patronis and referred the same day to the House Committee on Energy and Commerce. Both bills are short, the House version just six pages, and both simply repeal Section 230 effective two years after enactment, without proposing any replacement liability framework.

Other bills take narrower approaches. The Big Tech Accountability Act (H.R. 10300) aims to foster accountability for digital content providers, and the PROTECT Act (H.R. 7045) would repeal Section 230 protections in specified contexts. As of October 2026, none of these has become law; all sit at the introduced-and-referred stage.

Worth noting as context: Congress did pass a major adjacent law, the TAKE IT DOWN Act, signed on 19 May 2025 as Public Law 119-12. It makes it a federal crime to knowingly distribute nonconsensual intimate imagery, real or AI-generated, and requires covered platforms to remove reported imagery within 48 hours of a valid request, with FTC enforcement of the platform duties beginning in May 2026. It shows the current legislative pattern: rather than rewriting Section 230 itself, Congress keeps legislating around it, carving out specific harms while the broader immunity, and the broader fight, continues.

References

47 U.S.C. 230 (Communications Act of 1934); Congress.gov bill texts for S.3546, H.R. 10332, H.R. 10300, H.R. 7045, S.146; TechPolicy.Press September 2026 US tech policy roundup; Reason Section 230 Senate coverage; Federal Trade Commission TAKE IT DOWN Act enforcement guidance.