OpenAI faces concerns over teen ChatGPT use and mathematical findings

New research from Common Sense Media's Youth AI Safety Institute has designated OpenAI's ChatGPT for Teens an "unacceptable risk" for individuals under 18 years old. The report, updated on October 7, 2026, directly challenges OpenAI's assertions of enhanced safety protections for younger users, which were introduced with the ChatGPT for Teens launch in August 2026. Despite promises of parental notifications for concerning conversations, researchers found that these critical alerts frequently failed to materialize during simulated high-risk interactions.

The comprehensive study, involving over 4,000 prompts, revealed significant shortcomings in the chatbot's safeguards. Testers using newly linked teen accounts were able to engage in prolonged discussions about suicide, self-harm, and eating disorders, sometimes for up to an hour, without triggering any parental notifications. Furthermore, the research indicated a decrease in appropriate crisis referrals after the teen-specific features were implemented, and the "study mode" designed to assist learning could be easily circumvented by users seeking direct answers.

OpenAI has disputed the findings, suggesting that the testing methodology might not accurately reflect the full functionality of their teen-safety mechanisms or that some tests may have occurred before parental controls were fully activated. In light of the report, Common Sense Media's Youth AI Safety Institute is urging OpenAI to suspend the marketing of ChatGPT for Teens and restrict access for minors until independent verification confirms that adequate and effective protections are reliably in place.