Watchdog finds ChatGPT for Teens unreliable as OpenAI disputes results
Common Sense Media's testing of the teen version found weak safeguards and parent alerts, with crisis referrals dropping across 4,000 prompts. OpenAI questions whether the controls had actually switched on.
By Standard Topic. AI-generated from the reporting listed under Coverage. Published . Updated .
This story is no longer being updated. It reflects reporting up to .
What happened
Common Sense Media tested the teen version of ChatGPT and rated it an unacceptable risk. Its testing found crisis referrals dropped and Study Mode failed, and it reported unreliable safeguards and parent alerts. OpenAI disputed the findings, challenging whether the controls had been activated.
Why it matters
According to the testing, crisis referrals and parent alerts did not work reliably for teen users.
More from Technology & AI
- Eight states sue OpenAI, alleging misleading AI safety claims
- OpenAI announces solution to Navier-Stokes math problem
- Datadog bets AI agents can handle testing of AI-written code
- Anthropic and OpenAI sketch wealth-sharing plans before IPOs
- Mamdani administration proposes AI customer service rules
- Florida engineers study how AI behaves in high-risk situations