Meta uses AI profiling to infer user age, enforce teen restrictions

Meta has started using AI to infer user age and enforce restrictions for teens on its platforms. The system, detailed in a blog post, scans Instagram accounts in Brazil and the EU, as well as Facebook accounts in the U.S., for signals like birthday celebrations or school-related content to identify underage users. If the AI flags an account as potentially belonging to someone under 13, it will be deactivated unless the user provides age verification. The company claims this move aligns with its decade-long efforts to protect minors, including features like Teen Accounts with limited contact and content access. However, Meta’s track record contradicts this, as Australia’s data shows poor compliance with age laws, and U.S. courts have criticized its platforms for being designed to addict young users. Meta’s AI relies on profiling rather than third-party age verification tools like Yoti, which it has previously integrated. The system analyzes posts, comments, bios, and captions across Instagram Reels, Live streams, and Facebook Groups to detect underage accounts. Users flagged by the AI must verify their age to regain access, though Meta acknowledges the process is imperfect. Critics argue Meta’s approach prioritizes data collection over privacy, using age assurance as justification to expand AI access to user profiles. The UK Information Commissioner’s Office has classified this method as profiling, raising concerns about how personal data is used to enforce age restrictions. Meta also claims to support parents by providing guidance on discussing age honesty with teens, framing its efforts as part of broader online safety initiatives. Despite these assurances, the company’s history of non-compliance and design choices that attract underage users undermines its credibility in this area.