ChatGPT Rated 'Unacceptable Risk' for Minors After Parental Safety Alerts Fail in Testing

Over 4,000 test prompts conducted by the Common Sense Media Youth AI Safety Institute revealed that explicit conversations concerning suicide and eating disorders completely failed to trigger alerts on linked parent accounts. More than 25% of simulated crisis situations failed to point teenage users toward professional mental health resources or crisis referrals. Internal company data reportedly indicates that roughly two million people per week experience psychological harm from utilizing the service. The platform's behavioral age prediction system failed to identify minors using adult accounts, and test accounts explicitly stating they were 13 years old never automatically switched into teen safety modes. The complete failure of the Teen Safety Blueprint during independent testing directly bolsters ongoing lawsuits and regulatory actions, including Florida's active court effort to bar ChatGPT from minors entirely.