A new assessment from Common Sense Media's Youth AI Safety Institute has concluded that OpenAI's ChatGPT for Teens poses an "unacceptable risk" to adolescent users, prompting a call for the company to suspend marketing the product to teenagers until safety measures are substantially improved.
The institute conducted a comprehensive evaluation of the teen-focused version of the chatbot, which launched in August and is designed for users aged 13 to 17. Researchers ran more than 4,000 prompts through accounts configured to represent teenagers, testing the platform's safeguards across a range of mental health and educational scenarios.
Safety gaps in crisis response
The study found that while some protective features functioned as intended, several critical safeguards failed. Notably, the chatbot did not trigger parental alerts during prolonged conversations in which users explicitly discussed suicidal thoughts or self-harm. The report also highlighted deficiencies in how the system handled discussions related to eating disorders.
According to the assessment, ChatGPT for Teens missed more than one in four warranted crisis referrals across the mental health conditions tested. The platform also fell below the institute's 95% safety threshold on three of five severe-harm areas classified as "Red Lines."
Educational safeguards under scrutiny
Beyond mental health, the report questioned the effectiveness of educational features. OpenAI has stated that the teen version includes a Study Mode designed to guide students through problems with questions rather than direct answers. However, Common Sense Media found that the chatbot still allowed students to bypass these tutoring features and access answers directly.
The evaluation also identified issues with age verification. Some accounts registered as adults were not transitioned into the teen experience even after the users explicitly stated they were under 18, raising concerns about the platform's ability to enforce its age-based protections.
OpenAI disputes methodology
OpenAI responded by criticizing the testing approach, suggesting that much of the evaluation may have been conducted before parental controls were fully activated. The company stated, "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."
Common Sense Media rejected this explanation, asserting that it had confirmed with OpenAI that several teen-safety features were operational before testing began. The organization acknowledged that some parental-notification systems can take hours to activate but noted that some test accounts remained without alerts even after that period.
The report concludes with a direct appeal: "We call on OpenAI to pause marketing the product and to keep teens off of ChatGPT until it can offer a safe, developmentally appropriate experience."
This development comes amid broader scrutiny of AI safety for minors, with regulators and advocacy groups increasingly focusing on how emerging technologies handle vulnerable populations.
This article is for informational purposes only and does not constitute financial advice.
