On September 10, 2026, Governor Gavin Newsom signed AB 1709 (Lowenthal). It bars platforms from giving users under 16 features such as autoplay and algorithmic feeds based on user history and profile. It was one of thirteen child-safety bills signed the same day. The package also includes SB 1119, the companion-chatbot rules known as "Adam's Law", statutory damages of up to $1 million per child against large platforms that injure minors in AB 2, and an AI toy moratorium in SB 867.
The strongest case for the law
The case for AB 1709 deserves a fair hearing. Autoplay, infinite scroll and history-driven feeds are product design choices. They are built to lengthen sessions, and children are the users least able to resist them. The bill does not ban speech or any category of content. It regulates the delivery mechanism, which is a more defensible target than moderating what people say. Supporters such as Common Sense Media's Jim Steyer, quoted in the Governor's roundup, argue that "experimenting on children can no longer be the business model." The bill also follows California's 2024 SB 976, which targeted addictive feeds for minors, so it is a continuation of state policy and not a sudden shift.
If the law did only this, a proportionate-regulation view could accept it. The trouble is in how it is enforced.
The enforcement mechanism does the real work
Reporting on the law describes civil penalties of up to $50,000 per affected minor for knowing violations, enforceable by the Attorney General or local prosecutors. Those obligations begin January 1, 2027. A platform cannot know who is under 16 without knowing who is not, so the rule reaches every user. Tech Times' analysis notes that the penalty structure creates "strong commercial pressure on platforms to verify rather than trust" self-reported ages.
The state's age-signal infrastructure is the Digital Age Assurance Act (AB 1043), signed in October 2025. It has operating system providers collect a birth date at device setup and pass an age-bracket signal (under 13, 13-15, 16-17, 18+) to apps. That system also switches on January 1, 2027, the same day as AB 1709's platform duties.
The two laws pull in opposite directions. AB 1043 is a light-touch signal design. It has a known weakness: a 15-year-old who enters an adult birth date sends an 18+ signal to every app. Faced with $50,000-per-minor exposure, a rational platform will not rely on that signal alone. It will layer on facial estimation, ID checks or other verification that the statute never specifically asked for. Regulation by liability threat produces the most invasive compliance option, even where the text does not mandate it.
Speech costs fall on the young and on adults
The Electronic Frontier Foundation, which opposed AB 1709, calls it a functional ban on social media for under-16s and argues it violates young people's First Amendment rights. The EFF stresses that young people use these services for civic engagement, education, self-expression and community, and that isolated or vulnerable teens often have no substitute. Where a platform cannot separate feeds and autoplay for one age group, its practical options are to rebuild the product for minors or to exclude them.
The cost to adults is easy to overlook. Universal age assurance means every Californian hands identity or biometric data to a growing number of intermediaries. Data breaches, and the chilling effect on anonymous speech, are real costs. They also weigh against a law whose direct harm reduction has not been shown.
What proportionate regulation would look like
Two alternatives fit the evidence better.
First, regulate the design features and set defaults. Turn off autoplay and engagement-based ranking by default for minors, with a chronological or user-selected feed available. That reaches the same mechanism without forcing an identity check on everyone, because the burden falls on product defaults and not on gatekeeping.
Second, invest in the literacy approach the same package contains. The EFF supported two bills signed the same day, AB 2071 (digital wellness instruction in public schools) and AB 2298 (cybersecurity concepts in recommended curricula), and argues that open conversations about technology are far more effective than prohibition. That is an argument about relative effectiveness, and it is worth testing with evaluation requirements attached to these programs.
The package also creates an independent five-member e-Safety Advisory Commission within the Department of Justice, reporting annually to the Legislature. That is the right instrument if it is used well. Its first job should be to measure whether the ban reduces harm, and what it costs in verification data collected and teens shut out, before the state expands the model.
The larger risk
California is setting a template that other states will copy, and Hillary Clinton's endorsement in the Governor's roundup says as much: "More states should follow their lead." If the template is a liability-driven age-gate, then the internet's default for lawful speech becomes "prove who you are first." A regime that regulates the design of engagement features can protect children without that outcome. AB 1709 as written leans toward the more intrusive version, and the courts and the new commission should be watched closely before January 1, 2027.