Meta to delete accounts under 13

Meta is tightening age checks on Facebook and Instagram, and the new part is more aggressive than the usual “enter your birthday” gate. The company said on May 5 that it’s expanding AI systems that look for signs an account belongs to someone under 13, then deactivating those accounts unless the user can prove they’re old enough. The big shift is what Meta now counts as a clue. It already used text and behavior signals — things like birthday posts, school-grade references, bios, captions, comments, and other activity across a profile. Now it’s adding visual analysis too, which means AI can scan photos and videos for broad age cues like height or bone structure. Meta says that is not facial recognition and that the system is estimating general age, not identifying a specific person. Why does Meta care so much about 13? Because 13 is the floor in Meta’s own rules for Facebook and Instagram in most places. Kids younger than that are not supposed to have accounts at all. So this is not about moving a 15-year-old into a safer teen setting. It’s about finding accounts that, in Meta’s view, should not exist on the platforms in the first place. What happens if the system flags you? Meta says the account gets deactivated first. Then the user has to go through an age-verification process to show they are 13 or older. If they can’t, the account gets deleted. That matters because this is no longer just a recommendation engine quietly guessing your age in the background — it can directly lock you out. So why roll this out now? The timing is not random. Just days earlier, on April 29, the EU said Meta’s existing controls were failing to keep under-13s off Facebook and Instagram and warned the company it could face a fine of up to 6 percent of global annual turnover if those findings hold. Regulators said children could bypass the rules by entering fake birthdays and that Meta’s reporting tools were weak and hard to use. Meta is also pairing the under-13 crackdown with a broader teen-safety push. The company said it is expanding automatic Teen Account protections on Instagram in the EU and Brazil and on Facebook in the US for users it believes may actually be teens, even if they claimed to be adults. Basically, Meta is trying to sort users into two buckets with AI — under 13 gets removed, teens get stricter settings. The obvious catch is false positives. Age estimation is messy, especially from photos, video, slang, and social context. A young-looking adult, a short teenager, or a joke about being in middle school could all become noisy signals. Meta is clearly betting that an appeals-and-verification step is enough backstop, but that still means some legitimate users may have to prove their age to get their accounts back. That concern is an inference from how the system works, not something Meta spells out directly. The bottom line is simple. Meta has moved from asking users how old they are to actively guessing — and acting on the guess. For parents and regulators, that will look like overdue enforcement. For everyone else, it’s another example of platforms using AI not just to recommend content, but to decide who gets to be there at all.