Thirty Years, 26 Words, One Deadline
In 1996, Congress tucked 26 words into a telecommunications bill that would define the architecture of the internet: "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." Section 230 of the Communications Decency Act created the liability shield that let social media platforms, blogs, review sites, and online forums host user content without fear of becoming defendant in every defamatory post, harassment campaign, or dangerous video uploaded by their users.
Thirty years later, that shield is under simultaneous attack from two directions: in the courts, where judges have found a workaround; and in Congress, where a December 31, 2026 sunset deadline is forcing a long-deferred reckoning.
The Verdict That Went Around the Wall
On March 24, 2026, a Los Angeles jury did what no plaintiff had managed before in the mass social-media addiction litigation: it found Meta and YouTube liable — not for the content their users posted, but for the design of their platforms. The jury awarded approximately $6 million in damages in K.G.M. v. Meta Platforms et al., concluding that features like infinite scroll, autoplay, algorithmic recommendation, and beauty filters were defectively engineered to drive compulsive use, particularly in minors.
The Section 230 defense never had a chance. In a key pre-trial ruling, the court held that the immunity statute "did not bar plaintiffs' design claims" because the alleged harm flowed from the company's own engineering choices, not from anything a third-party user had written or posted. As American University's Kogod School of Business professor Erik Cogburn framed it: "If the platform itself is a defective product — think of a car with bad brakes — then Section 230 doesn't really come into play."
On April 10, 2026, a Massachusetts appeals court reached the same conclusion, ruling that unfair business practice claims targeting "addictive" design and inadequate age-gating were not covered by Section 230. Two rulings in two jurisdictions in six weeks amounts to a playbook — one attorneys across the thousands of pending cases in MDL 3047 are already adapting.
Congress: Inaction With an Expiration Date
Judicial pressure is arriving precisely as Congress has forced its own hand. On March 19, 2026, the Senate Commerce Committee convened a hearing titled "30 Years of Section 230," chaired by Sen. Ted Cruz (R-Texas). The hearing surfaced the familiar bipartisan diagnosis — the law needs updating — alongside familiar bipartisan disagreement about the cure.
Rep. Harriet Hageman (R-WY) has introduced H.R.6746, the Sunset to Reform Section 230 Act, which would strip the statute of legal force on December 31, 2026 unless Congress acts. A Senate companion, S.3546, mirrors that deadline. The theory is that a hard sunset creates legislative urgency where goodwill has failed. Whether that leverage produces thoughtful reform or a panicked repeal is the operative question.
Sen. Cruz, tellingly, has positioned himself against full repeal. At the hearing, he stated he was "concerned that a full repeal or sunset would lead platforms to engage in worse behavior" — meaning more preemptive censorship, not less. His preferred vehicles are the TERMS Act, which would restrict platforms from weaponizing terms-of-service to deplatform users, and the JAWBONE Act, targeting government pressure on platforms to remove content.
The TAKE IT DOWN Act, signed into law in May 2025, represents the only successful federal reform since FOSTA-SESTA in 2018. It criminalizes non-consensual intimate imagery including AI deepfakes and requires platforms to remove reported content within 48 hours. Crucially, it does not amend Section 230 directly — instead using FTC unfair-practices enforcement as a lever — demonstrating that targeted carve-outs can move through Congress even when comprehensive reform cannot.
The Strongest Case for Reform
Reformers deserve a fair hearing. Section 230 was designed for a 1996 internet of 40 million users and a few hundred thousand static websites, not the 4-billion-user algorithmic attention economy of 2026. When the law was written, platforms were passive hosts. Today, recommendation engines make active, revenue-optimized choices about what billions of people see next. Matthew Bergman of the Social Media Victims Law Center put it plainly at the Senate hearing: design decisions that addict young users to platforms "have nothing to do with protecting speech."
There is genuine consensus on at least one frontier: generative AI. Daphne Keller of Stanford Law, among the most credible voices defending Section 230, has acknowledged that AI-generated content presents a different case — the platform is the creator, not a neutral host. Justice Gorsuch flagged the same gap. Excluding AI outputs from Section 230 coverage has more bipartisan support than any other specific reform proposal.
Why Wholesale Repeal Would Backfire
Keller's broader warning is equally important: eliminating Section 230 outright would "very likely make the internet worse for user speech rights" and disproportionately harm smaller platforms that cannot absorb the legal costs large incumbents can. The Electronic Frontier Foundation has long documented that Section 230 protects not just Meta and YouTube but Wikipedia, Yelp, Reddit, independent blogs, and small community forums. A blanket repeal would accelerate consolidation, giving the largest platforms — which can afford litigation — a structural advantage over any challenger that emerges.
Nadine Farid Johnson of the Knight First Amendment Institute offered the most constructive framing at the March hearing: rather than removing Section 230 immunity, condition its availability on compliance with transparency, privacy, and interoperability standards. Platforms that want the shield would have to earn it through auditable behavior, not merely claim it.
The Right Reform Path
The December 2026 deadline should concentrate minds, not panic them. Congress has real options short of a blunt sunset: excluding AI-generated content from immunity; codifying the design-liability distinction that courts have now established twice; conditioning protection on meaningful transparency disclosures about algorithmic systems; and protecting researchers who need platform access to evaluate harms.
What Congress should resist is the temptation to treat "sunset" as a policy position. Letting Section 230 expire without a replacement — as H.R.6746 would do — does not return the internet to some pre-platform Eden. It floods courts with content-liability claims that judges are poorly equipped to resolve, pushes every medium-sized platform toward either over-moderation or closure, and hands the current incumbents a moat they could only have dreamed of in 1996.