According to a report released Wednesday by the Youth AI Safety Institute, a division of the California-based nonprofit Common Sense Media, OpenAI's ChatGPT for Teens, launched in August, poses excessive risks to users aged 13 to 17 and has been given an "unacceptable risk" rating.

This product is the default mode OpenAI has set for teenagers using its popular chatbot.

The report states that during testing, when teen accounts engaged in prolonged worrying conversations involving suicide, self-harm, or eating disorders, ChatGPT for Teens often failed to notify parents in a timely manner, or not at all; when these high-risk topics were involved, it also did not reliably suggest contacting crisis hotlines or seeking professional help.

The organization is calling on OpenAI to pause marketing of the product and to keep teenagers away from ChatGPT until it can provide a safe, age-appropriate experience.

The report notes that ChatGPT for Teens appears to notify parents only after an account has "tested hundreds of prompts over several weeks, covering more than a dozen conditions," rather than being triggered by a single concerning conversation.

The report also points out that ChatGPT cannot reliably identify teen users: when testers registered accounts claiming to be adults, even after subsequently telling the chatbot they were 13 years old, the account did not automatically switch to teen mode.

Common Sense is demanding that OpenAI improve its age estimation and parent alert mechanisms.

An OpenAI spokesperson said the company welcomes rigorous independent evaluations but does not believe Common Sense's testing accurately reflects how ChatGPT's teen protections actually operate, nor does it represent expert views on how AI can support teenagers.

The spokesperson said that after reviewing the report's methodology, OpenAI found that most of the testing may have begun and ended before OpenAI completed the rollout of teen parental controls; if so, the conclusions are inaccurate and cannot show whether parental safety notifications work as designed.

OpenAI said it has asked the nonprofit to re-test and emphasized its "deep commitment to teen safety," developing protections and giving parents meaningful tools.

Robbie Torney, head of AI and digital assessment at the Youth AI Safety Institute, said the organization stands by its testing methodology and results.

He said that before testing, it had confirmed with OpenAI that teen-related features such as eating disorder notifications had already been launched.

Regarding the parental notification feature, OpenAI informed them after testing was completed that activation on newly linked accounts could take several hours; some test accounts were linked within that window, while others were linked "significantly longer," yet still did not receive adult notifications.

Torney said the new information does not change the conclusion: in crisis situations, parent alerts are unreliable.

OpenAI has more than 1.2 billion ChatGPT users.

The company launched ChatGPT for Teens to make the chatbot safer for young people and added more parental controls, encouraging healthy AI use and critical thinking.

This move is also part of OpenAI's response to external scrutiny over its products' impact on teen mental health and cognitive development.

Previously, the parents of a 16-year-old user sued OpenAI, alleging its technology helped plan their son's suicide.

Common Sense Media stated that its funding sources include charitable organizations and tech companies, with the OpenAI Foundation being one of its donors, but the organization says it maintains editorial independence over its published content.

This controversy highlights that OpenAI, while expanding its teen user market, is facing rising AI safety and regulatory risks.

Common Sense's "unacceptable risk" rating for ChatGPT for Teens, along with its call to pause marketing, could increase pressure on OpenAI from parents, regulators, and the public.