Character.AI ends open-ended chat for under-18 users
Teen users lost open-ended chatbot conversation entirely, kept to two hours a day and shrinking until the cutoff, with an age-verification model built with Persona replacing it.
- Culture & impact
- Courts & copyright
- Notable
Character.AI announced it would eliminate open-ended chat with its AI characters for users under 18 by 25 November 2025, replacing it for that age group with alternative features such as video creation, story writing and streaming with characters that do not involve free-form conversation. In the run-up to the cutoff, teen chat time was capped at two hours a day and set to shrink further before the deadline. The company said it would deploy an in-house age-verification model alongside a third-party service, Persona, to determine which accounts the restriction applied to.
The announcement cited “recent news reports” and questions raised by regulators, safety experts and parents about how open-ended AI chat might affect teenagers “even when content controls work perfectly” — language that stopped short of naming the litigation Character.AI was facing. The company was sued in 2024 over the death of a Florida teenager who had formed an intense attachment to one of its chatbots, and three further families sued Character.AI and Google the previous month alleging its chatbots had isolated and sexually engaged with minors, including one 13-year-old who died.
Alongside the product change, Character.AI said it was setting up an independent nonprofit, the AI Safety Lab, to fund safety research on AI companionship products and inviting other technology companies, academics and policymakers to take part. The move made Character.AI the first major AI-companion platform to withdraw open-ended conversation from minors outright rather than relying solely on content filtering, after more than a year of lawsuits and state attorney-general scrutiny over chatbot harms to teenagers.