A nonprofit group that evaluates technology and media for families, Common Sense Media, published a safety assessment through its Youth AI Safety Institute that rated a teen focused version of ChatGPT an 'Unacceptable Risk.' The rating followed more than 4,000 test prompts designed to probe how the product responds in sensitive situations, one of the more extensive independent reviews of a mainstream AI chatbot's teen safety features to date.
The assessment's most pointed finding involved parental notifications. ChatGPT for Teens includes an option for parent linked accounts, intended to give parents visibility into how their teenagers are using the product. According to the testing, parent linked accounts received zero notifications during simulated hour long conversations that involved self harm and suicide related content, meaning the safeguard did not flag the kind of conversation it appears designed to catch, even over an extended, clearly sensitive exchange.
The testing also found that referrals to crisis hotlines, prompts meant to connect a user in distress with outside help, became less consistent over the course of the evaluation. Hotline referral rates on crisis related prompts dropped from 33% to 23%, suggesting the product's responses to the most sensitive queries were not uniformly reliable throughout the testing process, and that the system's behavior in this area can vary meaningfully from one conversation to the next.
Assessments like this one are part of a broader, ongoing conversation about how AI chatbots should handle young users and sensitive mental health topics, an area regulators, parents, and AI companies themselves have all been paying closer attention to as these products become more widely used by teenagers. Ratings from independent groups like Common Sense Media are meant to give parents and policymakers an outside benchmark, separate from what AI companies say about their own safety systems, and they often become a reference point in wider discussions about appropriate safeguards for younger users.
Watch for a response detailing any changes to the product's safety systems, and for whether other independent evaluators publish similar testing in the coming weeks.
What is Common Sense Media?
It is a nonprofit organization that rates and reviews media and technology products for families, and it runs a dedicated Youth AI Safety Institute focused on evaluating AI products used by children and teens.
What does the parent linked account feature in ChatGPT for Teens do?
It is designed to give parents visibility into their teenager's use of the product, including alerts in certain sensitive situations, though the testing found gaps in how consistently those alerts were triggered.