OpenAI Launches ChatGPT for Teens with Enhanced Safety Measures
OpenAI has introduced a version of ChatGPT designed specifically for teenagers, aiming to provide an age-appropriate and safe environment for users aged 13 to 17. This new iteration focuses on maintaining stronger protections, such as restricted content related to self-harm, suicide, and romantic engagements. The chatbot also offers study support, encouraging students to learn actively instead of merely providing answers.
Ann O’Leary, vice president of global policy at OpenAI, stated that the goal is to interact with teens in a manner suitable for their developmental stage. The company seeks to avoid exposing them to inappropriate content while not talking down to them or treating them like children.
“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,” said Ann O’Leary.
Brendan Bouffard, associate general counsel for AI and privacy at Fairplay, expressed cautious optimism about these changes. He emphasized the need to address potential emotional dependence on AI by young users. While the initiative is seen as progress, Bouffard warned against complacency regarding safety based on announcements alone.
Despite previous concerns over ChatGPT’s potential for misuse, especially related to facilitating cheating and unhealthy behaviors, OpenAI is working to mitigate risks. A 2025 study from Common Sense Media notes that 70% of teens use AI chatbots for companionship, with half engaging regularly. OpenAI CEO Sam Altman acknowledged the issue of emotional overreliance among youth.
The new ChatGPT for Teens version blocks any suggestions that the chatbot has personal feelings or consciousness, and prevents romantic or intimate dialogues. Allison Mishkin, head of child development at OpenAI, indicated that the team identified possible cues that might lead to users forming attachments to the chatbot.
“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,” noted Mishkin.
OpenAI does not verify users’ ages but uses age assurance techniques to identify minors based on query patterns. If a user is flagged as under 18, they are automatically placed into the teen-specific chatbot version. This system mirrors Meta’s strategy for Instagram, which enforces stricter rules for teen accounts.
Parents have the option to enable additional controls, allowing them to set ‘quiet hours’ and receive alerts in potentially risky situations. OpenAI mentioned additional notifications related to eating disorders, emphasizing the importance of offline support during crucial times.
Bouffard highlighted some safety gaps, such as the lack of automatic restrictions on ChatGPT’s long-term memory, which can only be addressed through parental controls. For educational support, the teen chatbot guides students to derive answers independently, aligning with OpenAI’s broader educational initiatives. The company is also investing in interactive learning methods to enhance educational efficacy.