The San Francisco-based outfit introduced ChatGPT for Teens aimed at children aged 13 to 17 who already use AI for schoolwork, daily questions and companionship.
OpenAI says the teen version comes with stronger protections, including restrictions around suicide, self-harm and romantic or sexual chats. It claims the system will support homework and study without simply belching out answers and essays.
According to The Associated Press the big idea is to steer teenagers towards healthier AI use in an age-appropriate space, the company said.
OpenAI vice-president of global policy Ann O’Leary said: “We want to treat teens like teens, which means that we have to make sure that we’re showing up with the right developmental stage when we’re not either talking down to them or treating them like kids, but we’re also making sure that they’re not exposed to material that they shouldn’t be exposed to.”
Whether the guardrails work is another matter. Fairplay associate general counsel for AI and privacy, Brendan Bouffard, said it was unclear whether OpenAI had done enough to address compulsive use and emotional dependence on chatbots.
“Announcements like these can be seen as a step in the right direction. But I also would caution against parents being lured into this false sense of safety based on the announcement,” Bouffard said.
Parents, teachers and child development experts have been sounding the alarm about children using AI chatbots. The machines have already been blamed for helping pupils cheat and, in darker cases, becoming tangled up in suicide risks.
Adults can be suckered into treating AI as a person, but teenagers are a softer target. Their brains are still developing, which makes the whole chatbot pal act a bit more worrying than the usual Silicon Valley nonsense.
Last year, watchdog research found ChatGPT would tell 13-year-olds how to get drunk and high, hide eating disorders and write a suicide letter to their parents. The chatbot usually warned against risky behaviour, then dished out detailed plans for drug use, calorie-restricted diets or self-injury.
In the US, more than 70 per cent of teenagers use AI chatbots for companionship, according to a 2025 study by Common Sense Media. Half use AI companions regularly, which should make every school and parent twitchy.
OpenAI CEO Sam Altman has said the company is studying “emotional overreliance” on the technology. Last year he described it as a “really common thing” with young people.
For ChatGPT for Teens, OpenAI says the bot is blocked from suggesting it has personal feelings or implying it is conscious. Romantic and sexual chats are meant to be blocked too.
OpenAI head of child development Allison Mishkin said: “We went through and identified what are the hypothetical cues that a model could give that might make a teenager kind of develop a relationship to it.”
OpenAI does not verify users’ ages. It says it uses age assurance to estimate whether someone is less than 18 based on signals such as their queries.
Users identified as minors or who identify as minors are automatically placed in the teen version. The approach resembles Meta’s teen accounts on Instagram, which have tighter content, chat and privacy settings.
Parental controls require the teenager and the parent or guardian to opt in. Parents with linked accounts can set quiet hours and receive safety alerts in limited high-risk cases, including possible self-harm.
“We are adding additional notifications related to eating disorders, while limiting what is shared and focusing on moments when offline support may matter most,” OpenAI said.
