OpenAI has updated ChatGPT’s guidelines to prioritise safety for users aged 13 to 17, even if it limits open-ended conversations.
ChatGPT will steer teens toward safer responses, encourage real-world support, and apply age-appropriate interaction standards.
Anthropic is developing age-detection tools to identify underage users and restrict or disable accounts where necessary.
OpenAI and Anthropic are now introducing new safety measures to make their AI chatbots safe for teenager use. This comes after the increased scrutiny from lawmakers and regulators over the impact of generative AI on young people’s mental health. OpenAI has updated its ChatGPT’s Model Spec, its internal guidelines that shape how the chatbot responds to users, to introduce specific safeguards for people aged 13 to 17. It will prioritise teen safety even in cases where it may limit other objectives, such as open-ended discussions or unrestricted exploration of ideas.
SurveyWith the new guidelines, ChatGPT is said to divert the younger users toward safer choices when conversation touches on sensitive or potentially harmful topics. The company has also stated that the chatbot will also place greater emphasis on encouraging real-world support, including offline relationships and trusted adults, and will set clearer boundaries when interacting with teens. The chatbot will now offer respectful and age-appropriate responses, avoiding both patronising language and adult-level framing.
These changes come from growing political and legal pressure on AI companies. OpenAI is currently facing a lawsuit that alleges ChatGPT provided harmful guidance related to self-harm to a minor. In response to broader concerns, the company has already introduced parental controls and restricted discussions around suicide for teenage users.
The company is also testing an age estimation system, which will attempt to identify whether a user may be under 18. If flagged, teen-specific safeguards would automatically apply, while adults incorrectly identified as minors would be able to verify their age.
Anthropic is also working on a similar approach. It is developing tools to detect subtle conversational cues that will indicate a user is underage. Accounts confirmed to belong to minors will be disabled, and the company already flags users who disclose their age during conversations.
Follow Us
Ashish Singh
Ashish Singh is the Chief Copy Editor at Digit. He's been wrangling tech jargon since 2020 (Times Internet, Jagran English '22). When not policing commas, he's likely fueling his gadget habit with coffee, strategising his next virtual race, or plotting a road trip to test the latest in-car tech. He speaks fluent Geek. View Full Profile
