Common Sense Media's Youth AI Safety Institute said that in its tests, teen accounts could hold lengthy conversations about suicide, self-harm, or eating disorders without parents being alerted quickly, if at all. The group added that in cases where teens brought up high-risk issues, the ChatGPT for Teens version often failed to direct them reliably to crisis hotlines or professional support.
Their testing spanned before and after the teen product's launch, using more than 4,000 prompts on accounts set up as 13 to 17 year olds, and the team shared ChatGPT's responses with child psychiatrists and a pediatrician. The nonprofit said alerts to parents tended to show up only after accounts had been run through hundreds of prompts across many scenarios over weeks, rather than after a single alarming exchange. The report further noted weak age recognition by the chatbot: accounts set up as adults weren't automatically moved to the teen experience, even when the user told the bot they were 13.
Common Sense gave ChatGPT for Teens an "unacceptable risk" label and asked OpenAI to improve both age detection and the rules for when to notify parents. The group said it shared its findings with the company before publishing.
OpenAI pushed back. "We welcome rigorous independent evaluation, but we do not believe Common Sense Media's testing accurately reflects how ChatGPT's teen safeguards work in practice or expert perspectives on how AI can support teens," a spokesperson said, adding, "We are deeply committed to teen safety and to developing safeguards and giving parents meaningful tools to guide their teens' use of AI."
The company said most of the testing may have started and ended before it finished rolling out parental controls for teens, which would skew the findings, and asked the nonprofit to run the tests again. OpenAI also told the group it can take several hours for parent-safety features to activate on newly linked accounts. On age detection, OpenAI said it uses multiple signals over time to route people to the right experience, and while stating an age in chat is one input, that alone does not reclassify an account.
Independent reviews of AI products are starting to shape what regulators demand. Market Briefs covers AI oversight free every weekday.
Robbie Torney, who leads AI and digital assessments at the Youth AI Safety Institute, said the nonprofit stands by its methods and results. He said the team confirmed with OpenAI before testing that teen-focused features, including eating disorder notifications, had been rolled out. Afterward, OpenAI told them about the activation lag on linked accounts.
Torney said certain test accounts were connected during that activation period and others for much longer spans, yet no adult notifications showed up. "This new information does not change our conclusion that parental alerts are unreliable for crisis situations," he said. Torney also added, "I think the burden is now on OpenAI to demonstrate that their product is safe for teens to use."
The organization called on OpenAI to halt promotion of the product and "to keep teens off ChatGPT until it can offer a safe, developmentally appropriate experience."
OpenAI, which counts more than 1.2 billion ChatGPT users, introduced the teen mode to promote safer use and build healthy habits for younger people. The rollout is part of broader changes under pressure about youth mental health and development, especially after parents of a 16 year old sued the company last year, alleging its technology aided in planning their son's suicide.
Common Sense tested before and after the teen release, asked experts to review the outputs, and flagged age estimation and alert timing as priority fixes. The nonprofit also discloses that it's funded by philanthropies and tech firms, including support from the OpenAI Foundation, and says it maintains editorial independence.
For your money, this is a reminder that products built for minors attract intense scrutiny. Ratings like "unacceptable risk," debates over safeguards, and calls to halt marketing can shape public perception and policy pressure. That can influence how fast features aimed at younger users roll out and how companies communicate safety standards, which in turn affects brand trust and long term adoption.
Safety assessments increasingly carry commercial consequences. Join Market Briefs free and follow the scrutiny.
