OpenAI is implementing stricter limits on the types of weapons-related information minors can access through ChatGPT. The new policy applies even when a question "appears harmless," aiming to prevent potential misuse. The AI company stated it already has safeguards against facilitating violent acts, but is now specifically restricting practical details about firearms and security systems for teenage users.
Teens will still be able to use the chatbot for violence prevention, de-escalation, and seeking help from trusted adults. These changes complement existing parental controls that notify guardians of policy violations. The updated rules also extend to all users, blocking detailed reconstructions of past mass attacks and comparisons of harm scales.
This move follows a wrongful death lawsuit and CEO Sam Altman's apology regarding the Tumbler Ridge shooting. OpenAI plans to release its first Community Safety Report in early 2027.
Source: Engadget · Summarized by HeadlinesBriefing