Age prediction tools, like those YouTube began implementing last year, are officially rolling out to ChatGPT, OpenAI announced in a blog post yesterday. The idea is to have the chatbot automatically detect when it’s chatting with a minor, and dump them in a separate pool that’s a little more sterile than a typical conversation.
“Age prediction builds on protections already in place. Teens who tell us they are under 18 when they sign up automatically receive additional safeguards to reduce exposure to sensitive or potentially harmful content. This also enables us to treat adults like adults and use our tools in the way that they want, within the bounds of safety,” it wrote.
ChatGPT is down with the kids
ChatGPT has been granted a far more scrutinous eye than ever before, allowing the model to judge the contents of its chats with users to determine whether they’re underage. It’ll look at a “combination of behavioral and account-level signals” like an account’s age, the times when a user is active, and most importantly, the user’s stated age.
Age prediction is a crucial step for ChatGPT. It allows OpenAI to turn its attention to the incoming “adult mode”, which Fidji Simo, the company’s CEO of applications, said was expected to debut inside ChatGPT in the first quarter of 2026. If the model detects it’s chatting with an underage user, it’ll cordon off the bot’s spicier elements.
Read More: ChatGPT Health wants all of your medical data stored in one (AI-filled) place
Those include “depictions of self-harm, graphic violence or gory content, sexual, romantic, or violent role play, content that promotes extreme beauty standards, unhealthy dieting, or body shaming, and viral challenges that could encourage risky or harmful behavior in minors.” Anyone incorrectly labelled as a minor will be required to upload their government ID (yeah, right) or present their face in the form of a live selfie, handled by Persona.
If a user isn’t bothered about losing access to, uh, “content that promotes extreme beauty standards,” it could be the new meta to allow ChatGPT to mark them as underage. Not only does it mean fewer invasions of privacy, but an underage user may successfully fend off ads by the time OpenAI is ready to roll them out everywhere.
