California has banned the algorithm for under-16s, not the account

California’s governor Gavin Newsom signedAssembly Bill 1709into law on 10 September 2026, barring social media platforms from giving any user under 16 an "addictive feature". The Act defines that as two things: a feed that recommends other people’s posts based on what the platform knows about you, and autoplay. It passed78 votes to nil in the Assembly and 39 to nil in the Senate.

The account itself survives. Section 22683 says a platform may still let an under-16 sign in, so long as that account carries no addictive feature. A platform that will not build the stripped-back version has to delete the child’s account and the personal information attached to it. Before serving a recommended feed to anyone, platforms must check the user’s age under California’s Digital Age Assurance Act, which starts on 1 January 2027.

Only the Attorney General or a local prosecutor can bring a case, and the penalty is counted per child: up to $50,000 for each affected under-16 where the breach was knowing, and up to $25,000 where it was negligent. The law also sets up an e-Safety Advisory Commission inside the Department of Justice, which has to report to the legislature every January. It advises and nothing more. Its recommendations bind no one.

Why this matters for a UK parent

Three countries are now drawing the same line in three different places, and the UK’s version is the only one a child can undo. From spring 2027 an under-16 here loses the account outright, while the measures aimed at compulsive scrolling (themidnight-to-6am curfew and autoplay switched off) apply to 16- and 17-year-olds and can be turned back on by the teenager. California leaves the account alone and makes the feed restriction compulsory. Australia’sdraft duty of care keeps its ban and hands everyone over 16 an off switch for the algorithm.