The Section 230 question that decides whether infinite scroll is a product defect

An appeal in the Meta litigation now turns on one narrow question: whether Section 230 of the Communications Decency Act bars state claims that Meta committed unfair and deceptive trade practices by knowingly implementing platform design features that harmed users for profit ^[1]. That question decides whether engagement mechanics are treated as publishing decisions, which Section 230 protects by barring courts from treating internet companies as the publishers of user-generated content ^[9], or as design decisions, which it does not reach.

The procedural history is worth keeping straight. In October 2023, forty-two attorneys general sued Meta, alleging that specific platform design choices drive compulsive use and that company statements deceived the public about Meta's design goals ^[2]. Thirty-three of those cases were folded into the multidistrict litigation in the Northern District of California titled In re Social Media Adolescent/Personal Injury Products Liability ^[3]. Nine remain outside that MDL ^[4]. The caption already says products liability; the appeal decides whether the theory survives contact with Section 230.

Meta moved to dismiss on the ground that the claims concerned its publishing activities ^[5]. The district judge split the difference, granting the motion in part and denying it in part, and dismissing the claims that targeted infinite scroll, autoplay, usage-maximization and similar features ^[6]. Meta appealed the denials; the states cross-appealed the dismissals ^[7]. Because the case sits in the Northern District of California, the appeal runs to the Ninth Circuit, whose Section 230 case law EPIC says has been progressively narrowed over the last fifteen years even as many district and appellate judges remain confused about how to apply it ^[13][12].

EPIC filed an amicus brief supporting the states, joined by Common Sense Media, Cybersecurity for Democracy, the Tech Justice Law Project and a group of law professors ^[8]. Its argument is historical: Section 230 grew out of early internet defamation cases where courts had to choose between treating companies as publishers, who have a common law duty to keep their publications free of unlawful content, or distributors, who do not ^[10]. Treating moderation as publishing created the moderator's dilemma, which the statute was written to solve ^[11]. On EPIC's reading, that has nothing to do with whether a scroll ends, and the district judge misunderstood both the statute's purpose and the circuit's precedent when he dismissed those claims ^[14].

For anyone who ships a feed, the exposure is not the content and never was. The features named in the complaint are the ordinary furniture of retention work: infinite scroll, autoplay, and usage-maximizing systems built into recommendation algorithms, which EPIC describes as causing people to use a platform more than they would like ^[15]. The deception half of the case runs on intent, and whistleblowers allege that executives knew about the harmful effects and repeatedly declined alternative designs that would have reduced profits while saying the opposite in public ^[16]. That is a claim about the decision record: experiment results, rejected variants, and the metric the variant was rejected for.

Watch how the court draws the line between a feature and the content it arranges, and whether the cross-appeal revives the infinite scroll and autoplay claims. Watch the nine attorney general cases outside the MDL ^[4], which give states a second venue if the appeal goes badly, and the state laws already being passed to regulate design features directly ^[17].