OpenAI adds stricter limits for teens asking about firearms
Reported by Reuters, Blockchain News and Engadget3 sources ↓
OpenAI says ChatGPT will restrict practical weapons and security information for teen users, even when a question appears harmless.
Key points
- OpenAI says teens can still ask for help preventing or de-escalating violence and connecting with a trusted adult
- The company is also tightening limits for all users on details about past mass attacks
- Existing parental controls can notify parents about linked teen accounts that violate policies on violent threats or acts
- OpenAI says its first Community Safety Report will come in early 2027, with updates every six months
Sources
Details
OpenAI said it will restrict practical information about weapons and security systems for teen users, including when there is no sign of violent intent. The company said teens can still use ChatGPT for help preventing or de-escalating violence, seeking support and connecting with a trusted adult.
OpenAI also said it is strengthening rules for all users on reconstructions of past mass attacks, including details about weapons and how they were obtained, and comparisons of the scale of harm. The new restrictions will operate alongside parental controls that can notify parents when linked teen accounts violate policies on violent threats or acts of violence.
OpenAI said it plans to publish its first Community Safety Report in early 2027 and issue updates every six months.
Written from coverage as of