Meta has announced it will deploy artificial intelligence to identify and remove accounts belonging to people under 13, the minimum age allowed on Facebook and Instagram. In an early May statement, the company said it will analyze whole profiles for contextual signals — birthday posts, mentions of school grades, bios, captions, comments and other content — to flag accounts that may belong to children who signed up with a false birth date.
The project, which Meta says will be powered by its own “Meta AI,” has drawn scrutiny because the technology will also examine images for physical cues such as height and bone structure. Critics call that practice invasive and warn it risks building extensive age-based profiles of young people.
Meta’s announcement came days after the European Commission published preliminary findings that the company had failed to prevent minors under 13 from using Instagram and Facebook in the EU. Meta says it does not use data from children under 13 to train its AI and that it will delete data from accounts found to belong to underage users. It has said accounts that the AI flags will be temporarily deactivated and removed permanently if the holder cannot prove they are over 13.
Experts and privacy advocates caution that automated age detection has serious pitfalls. Nina Kolleck, a professor of educational and socialization theory, notes that the AI will need extensive data to infer age and behavior — data that could amount to intrusive profiling. Andy Przybylski, a researcher of human behavior and technology, warned that collecting and processing young people’s faces and behaviors is unlikely to make them safer and could instead create a verified list of advertising targets.
Other critics argue that focusing on age verification sidesteps deeper platform harms. Stephan Dreyer, a media law and governance researcher, points to platform features that can damage adolescent development — endless feeds, recommendation systems that emphasize beauty ideals, and discriminatory content — and says legal remedies should address these aspects rather than only enforcing minimum ages.
Internationally, the question of who should be allowed on social media is contentious. Australia and Indonesia have recently introduced laws restricting access to social platforms for people under 16. Several EU countries, including Germany, France and Poland, are discussing similar measures. But many experts and young people push back against blanket bans. A representative April 2026 poll by UNICEF Germany found that 74% of 14- to 16-year-olds oppose a social media ban for those under 16. UNICEF officials note that social media is embedded in modern social life and that access can be especially important for young people from vulnerable, minority or mobile-challenged groups — for example refugees, LGBTQ+ youth, and people with disabilities.
Some researchers favour alternatives to outright prohibitions. Przybylski recommends stronger privacy protections to limit data collection from under-18s and better media literacy programs for young people and parents. He compares strict age cutoffs to “abstinence-only” approaches to sex education, arguing that learning to navigate online spaces safely is more effective than simple exclusion.
Parents express mixed feelings. One mother interviewed in Germany said she finds the idea of AI scanning profiles troubling because Meta is a “data hoarder” and she doubts the company will truly delete information collected about children. Meta maintains it will remove data from underage accounts and that flagged accounts must verify age or be deleted.
As tech companies and regulators debate how to limit minors’ exposure to social platforms, the policy trade-offs are clear: automated age-detection tools may help enforce rules on paper but raise privacy and profiling concerns; outright bans risk isolating young people; and targeted regulation of platform features and stronger privacy protections are being offered as alternative or complementary solutions. For now, Meta plans to roll out the AI-based checks and deactivate accounts it believes are underage unless users can prove otherwise.