A Bill Aimed at Minors, Built on Everyone's Data
California's legislature passed Assembly Bill 1709 on August 31, 2026, and enrolled it on September 4, sending it to Governor Gavin Newsom for signature or veto. Authored by Assemblymember Josh Lowenthal, the bill bars "covered platforms" from providing addictive features — a category that includes algorithmic recommendation feeds, autoplay, and infinite scroll — to any user under 16. If Newsom signs it, the restriction takes effect alongside California's separate Digital Age Assurance Act, which requires age-bracket signals starting January 1, 2027.
The steelman case for AB 1709 is not frivolous. Legislators and groups like Common Sense Media point to a genuine, well-documented problem: engagement-optimized feeds are designed to maximize time-on-app, and there is real clinical concern about compulsive use among adolescents. The bill's committee votes were lopsided — 13-1 in the Privacy and Consumer Protection Committee, 10-0 in Judiciary — reflecting a bipartisan judgment that self-regulation by platforms has not addressed the problem. A narrowly drawn rule limiting engagement-maximizing design for children is a defensible policy goal.
The problem is that AB 1709 is not narrowly drawn. To enforce an under-16 restriction, a platform must first know which users are under 16 — which means verifying the age of every user, adult and minor alike. As the Electronic Frontier Foundation detailed in its July 31 critique of the bill's amendments, compliance will functionally require platforms to collect government-issued ID or run biometric age-estimation scans on their entire user base, not just the population the law is trying to protect. EFF's August 31 letter urging a veto called the result "a sweeping ban on social media use for young people under the age of 16" that also imposes new privacy burdens on adults.
The Definition Problem
AB 1709's core definitional move is what makes it sweeping rather than surgical. The bill defines an "addictive feed" as any presentation of user-generated content recommended "in whole or in part" on information the user has provided — which, per EFF's analysis, covers virtually every functional recommendation algorithm, including basic signals like who a user follows or what posts they have liked. That is not a definition targeting manipulative dark patterns; it is a definition that captures ordinary content ranking. A platform cannot offer a working, browsable feed to a 15-year-old without triggering the statute, because almost any relevance ranking uses the same signals the bill labels "addictive."
The enforcement stakes are real: violations are civil, with penalties up to $50,000 per affected minor for knowing violations and $25,000 for negligent ones, enforced by the Attorney General or local prosecutors. Facing that exposure, platforms have no incentive to build a lighter-touch compliance path — they will verify age broadly and conservatively, which is exactly the outcome EFF warns creates "honeypots" of sensitive identity data sitting on private servers, an attractive target for breach regardless of how well-intentioned the underlying law was.
Courts Are Still Working Out the Rules
The legal terrain AB 1709 will face has shifted, and both sides can find something to like in it. In June 2025, the Supreme Court decided Free Speech Coalition v. Paxton, upholding Texas's age-verification requirement for adult content sites under intermediate — not strict — scrutiny, a standard significantly friendlier to state age-gating laws than what applied before. That ruling gives California's lawyers a stronger hand than they would have had two years ago.
But AB 1709 is a materially different animal from the Texas statute: it does not target a narrow category of obscene-to-minors content, it restricts core functionality of general-purpose speech platforms for an entire age cohort. The Ninth Circuit's March 2026 ruling in NetChoice v. Bonta — which narrowed but did not eliminate the injunction against California's Age-Appropriate Design Code Act — shows the harder scrutiny state legislators still face when a law reaches beyond narrow age-verification into broad platform-design mandates like data-use restrictions and "dark patterns" bans, which the panel found unconstitutionally vague. AB 1709's addictive-feed definition has the same vagueness problem the Ninth Circuit just flagged in a sibling statute.
The Better Path Was Available
California did not need to choose between doing nothing and mandating identity verification for every resident who wants to use social media. A design-code approach — banning specific manipulative mechanics (autoplay defaults, streaks, push-notification dark patterns) without conditioning access on ID checks — targets the same harm without the privacy externality. Newsom has vetoed sweeping tech bills before on proportionality grounds; AB 1709's core flaw isn't its goal, it's that the compliance mechanism punishes the 90-plus percent of users who are adults to reach the minority who are minors. A veto, paired with a narrower redraft focused on default settings rather than identity checks, would achieve more of what the bill's authors actually want.