In mid-September 2026, Reps. Mark DeSaulnier (D-CA) and Jimmy Patronis (R-FL) introduced the Sunset Section 230 Act, H.R. 10332. It would repeal Section 230 of the Communications Decency Act two years after enactment and put nothing in its place. The bill went to the House Energy and Commerce Committee. Its sponsors argue that "Section 230 was created in 1996 as part of the Communications Decency Act to help encourage innovation and growth among new social media companies, but 30 years later as these companies have become inordinately powerful, Section 230 has become obsolete and dangerous," according to a release summarized by the Benton Institute.
The strongest case for the bill
The sponsors' case deserves a fair hearing. Families of children harmed by sextortion, drug sales and cyberbullying often cannot get a court to consider whether a platform's own product design contributed. Immunity can end a case before discovery. Advocates such as attorney Carrie Goldberg and Marc Berkman of the Organization for Social Media Safety made this argument at a 2024 House Energy and Commerce hearing on a similar sunset proposal, according to a Tech Policy Press transcript. A sunset clock is also a forcing device. Congress has tried to amend the statute for years, and a deadline might finally make it write something better. That is a serious argument.
What the law actually does
Section 230(c)(1) says: "No provider or user of an interactive computer service shall be treated as the publisher or speaker of any information provided by another information content provider." The text is available in the U.S. Code via GovInfo. The statute is narrower than its critics often suggest. Section 230(e)(1) already leaves federal criminal law untouched, including the chapters on obscenity and child sexual exploitation. Intellectual property claims and the 2018 sex-trafficking carve-out also sit outside the shield. The immunity covers third-party speech. It does not cover a company's own conduct or its own content.
Repeal would therefore not mainly change how platforms handle the worst material, which federal criminal law already reaches. It would change who can be sued over ordinary speech: reviews, forum posts, comments, community discussion. The Supreme Court had a chance to narrow the statute in Gonzalez v. Google and did not take it. The Court issued its decision on May 18, 2023, and the question of how far Section 230 reaches stayed with Congress.
The design flaw: a repeal with no replacement
The central problem is structural. The bill removes a liability rule and puts no standard in its place. Platforms would not know what duty they owe, what notice triggers it, or what counts as reasonable moderation. Courts would work that out case by case over years, and companies would have to operate under that uncertainty in the meantime.
The predictable response is to remove anything that might draw a lawsuit. Engine's Kate Tummarello warned at the 2024 hearing that a sunset "risks leaving internet platforms, especially those run by startups, open to ruinous litigation," according to the same Tech Policy Press transcript. A platform facing liability for every user post has two options: pre-screen heavily or shut down comment features. Both reduce lawful speech. The Electronic Frontier Foundation notes that without Section 230 protection, many intermediaries would heavily censor or refuse to host user content. The same page points to Canada and Australia, where courts have held operators liable for user speech. EFF says that liability reduced the amount of user speech online, particularly on controversial subjects.
Who wins and who loses
Liability costs fall unevenly. A large platform can employ thousands of moderators, build classifiers and absorb litigation costs. A volunteer-run forum, a local news site with a comment section, or a niche review site cannot. Section 230 helps operators resolve meritless suits quickly, and without it even a winning defendant can be bankrupted by the cost of defending. A repeal that targets "big tech" in its rhetoric would in practice leave the biggest firms best placed to cope. That is the opposite of the competition outcome reformers say they want. It would entrench incumbents, whose legal budgets are the moat.
The bill's framing around children also does not match its scope. The sponsors emphasize protecting young people, but the repeal is universal. It applies to every blog, review platform and community server, not just to social apps used by minors. A narrower tool, such as a duty of care for specific design features or targeted exceptions like the 2018 sex-trafficking carve-out, can be argued on its merits and tested. A blanket repeal cannot be tested. It can only be survived.
A more proportionate path
If lawmakers want accountability, they have better options than a clock on the whole statute:
- Target conduct, not hosting. Liability for a platform's own design decisions or its own algorithmic amplification claims is a different question from liability for a user's post, and can be drafted separately.
- Create a notice-and-action standard with clear thresholds and safe harbors for good-faith compliance, so that operators know what the law expects of them.
- Tier obligations by size, so that compliance costs track a company's capacity to bear them.
- Hold hearings before setting a deadline. A sunset works as leverage only if a credible replacement is already drafted. Without one, the deadline simply becomes a date on which the rules disappear.
What to watch
The bill has been referred to committee, and committee referral is where most such bills stay. Similar sunset proposals were debated in 2024 without becoming law. Still, bipartisan sponsorship matters, and the same two-year idea has surfaced in the Senate. The right response from reform-minded lawmakers is to specify what should replace Section 230 before setting a date for it to expire. Without that, the likeliest outcome is not a more accountable internet. It is a smaller, more heavily filtered one, run by the few companies that can afford the risk.