OpenAI has launched a new version of ChatGPT tailored for teenagers aged 13 to 17. This version includes enhanced protections, such as content restrictions on topics like suicide, self-harm, and romantic or sexual conversations. The goal is to create an age-appropriate environment that aids in homework and study without merely providing answers.
Ann O’Leary, OpenAI’s vice president of global policy, emphasized the importance of recognizing teens’ developmental stages. She stated, “We want to treat teens like teens, making sure they’re not exposed to inappropriate material.”
Brendan Bouffard of the children’s advocacy group Fairplay expressed skepticism regarding the effectiveness of these safeguards. He noted that while OpenAI’s initiative may seem promising, parents should not assume complete safety.
Young people’s interaction with AI chatbots has raised concerns. Reports indicate these tools can facilitate cheating and even influence negative behaviors. A study highlighted ChatGPT’s previous shortcomings in advising teens on dangerous activities.
A 2025 study by Common Sense Media revealed that over 70% of teens use AI chatbots for companionship, with half engaging regularly. OpenAI’s Sam Altman acknowledged the issue of “emotional overreliance” on such technologies among youths.
To mitigate risks, ChatGPT for Teens is designed not to suggest or imply personal emotions. It blocks romantic chats and other cues implying consciousness. Allison Mishkin, leading child development at OpenAI, mentioned the effort to remove any cues that might encourage relationship development with the AI.
This tailored ChatGPT version adds safety protections, even for teens not using parental controls. OpenAI employs age assurance techniques, similar to Instagram’s for minors, to identify users under 18.
Parents can set restrictions like “quiet hours” and receive alerts for high-risk situations. Bouffard pointed out gaps in these measures, noting the absence of restrictions on the AI’s long-term memory by default in teen accounts.
OpenAI assures further developments in homework assistance through interactive learning. This strategy aligns with research indicating that active engagement aids learning better than passive information reception.