Meta has rolled out one of its most significant child safety upgrades to date, deploying artificial intelligence across Instagram and Facebook to identify and remove users under 13, and to automatically place teenagers into age-restricted account settings even when they have entered an adult birthdate to bypass restrictions.

The announcement, made on 5 May 2026, marks a significant escalation in how the world's largest social media company is approaching the problem of underage access on its platforms. Rather than relying solely on users to provide accurate age information during sign-up, Meta is now using AI to actively analyse account behaviour and content to infer whether a user is likely a minor.

How the Technology Works

Meta's AI systems scan a wide range of signals across user profiles, including posts, comments, bios, and captions, looking for contextual indicators of age such as references to school environments, birthday milestones, or age-related language patterns. The technology is being expanded across Instagram Reels, Instagram Live, and Facebook Groups to broaden its detection coverage.

A new visual analysis layer has also been added, allowing Meta's AI to scan photos and videos for visual cues about a user's age that text alone might miss. Meta was explicit that this system does not use facial recognition and does not identify individuals. It estimates age ranges based on broad visual characteristics, which when combined with behavioural signals, significantly improves detection accuracy.

When an account is flagged as potentially underage, it is deactivated and the holder must provide verified proof of age to prevent permanent deletion. Users who attempt to change their registered age in ways that may bypass protections face identity verification and facial age estimation requirements.



What This Means for African Users

For African parents and families navigating their children's social media use, the update carries direct relevance. Nigeria, South Africa, Kenya, and Ghana are among the continent's largest Instagram and Facebook markets, with millions of young users across all four countries. The gap between the minimum age requirement of 13 and the actual age at which many children first access these platforms has been a persistent concern for parents, educators, and regulators across Africa.

Meta's AI-driven approach removes some of the burden from parents and places it directly on the platform's own detection systems. Rather than depending on a child being honest about their age at sign-up, or on a parent monitoring their child's account activity closely enough to spot a problem, the platform is now taking proactive responsibility for enforcement.

The Broader Accountability Question

Meta's announcement comes as regulators across multiple markets intensify scrutiny of how social media platforms manage child safety. Kenya's government this week directed X to open a local office or face suspension, citing deepfakes and harmful content targeting children as primary concerns. The UK Online Safety Act, Australia's social media age restriction legislation, and similar frameworks across Europe are all pushing platforms toward more active enforcement rather than passive policy compliance.

Meta's AI age assurance rollout is a genuine step forward. Whether it is sufficient to meet the standards that regulators, parents, and child safety advocates are beginning to demand is a different and still open question.