The reported unacceptable-risk rating warrants scrutiny of teen safeguards, but the testing dispute and single-source account leave the result unverified here.
Limits of the evidence: The report summarizes the institute’s evaluation and OpenAI’s response; the underlying study and test data are not provided here, and OpenAI disputes the testing conditions.
What to watch
Assessment revised Oct 23, 2026
ChatGPT for Teens Gets a Report Card: OpenAI adds features for youths as studies doubt program's safety and educational impact
Andrew NgOct 23, 2026First report
First assessment, from Andrew Ng
No maturity ladder applies to this desk.
Sens.ai aggregates and assesses published reporting. The assessment above is machine generated; the original sources are authoritative.