Common Sense Media Report Deems ChatGPT Teen Mode Unsafe for Minors

The Youth AI Safety Institute at Common Sense Media published a report finding that ChatGPT's teen mode fails to reliably trigger parental alerts or recommend crisis hotlines during high-risk conversations like self-harm or suicide. The organization labeled the feature an "unacceptable risk" and urged OpenAI to pause its rollout to minors until safety controls are fixed. The findings highlight growing scrutiny over how AI models interact with minors and the technical challenges in automated age-gating and crisis intervention. It calls into question whether tech companies can adequately protect vulnerable young users from harmful content and psychological distress. Researchers found that triggering a parental notification required hundreds of prompts over several weeks regarding alarming topics, rather than a single acute conversation. Additionally, OpenAI disputed the report by claiming testing occurred before features were fully deployed, though researchers maintained OpenAI had explicitly confirmed system readiness prior to testing.