Meta Pauses AI Character Access for Teens: A Major 2026 Safety Pivot

Meta just pulled the plug on AI best friends. In a quiet 'Friday news dump,' the social media giant temporarily paused teen access to its AI characters globally—at least until the tech can learn to behave itself.

Meta AI Character Teen Access Pause Explained

The announcement came not as a fresh press release, but as an update to a blog post from October last year. According to the updated note from Adam Mosseri, Head of Instagram, and Alexandr Wang, Chief AI Officer, the company is building a 'new version' of AI characters. While this development is underway, teens will no longer be able to access existing personas. This restriction applies to verified teens and those suspected to be underage by Meta's age prediction technology.

This move is a clear admission of the risks surrounding AI companions and teen mental health. Meta isn't the only one facing scrutiny. This month, Character.AI and Google settled lawsuits involving the tragic suicides of children who had become obsessed with chatbots. In one case, a 14-year-old boy was reportedly groomed and abused by a bot based on a popular TV character. Despite Meta's earlier efforts to train bots to avoid topics like self-harm or inappropriate romance, those safeguards evidently weren't enough.

A Recurring Struggle for Responsible AI

This isn't Meta's first U-turn. In 2024, it scrapped celebrity-based AI personas, and in January last year, it pulled AI profiles following a racism controversy. The scale of usage makes this pause significant: more than half of teens aged 13-17 surveyed by Common Sense Media admitted to using AI companions at least once a month.