OpenAI tightens ChatGPT restrictions on weapons topics for teenagers
AI company OpenAI is introducing stricter limits on weapons-related information that teenagers can get from its ChatGPT chatbot. The limits will apply even to questions that look harmless on the surface.
OpenAI, the artificial intelligence company behind the ChatGPT chatbot, has announced stricter restrictions on the weapons-related information that teenage users can receive. The changes concern young people's access to content about firearms and are meant to keep such material away from minors.
What the new rules say
According to the company, the new rules block access to this type of content even when a teenager's question appears completely harmless and contains no obvious threat of violence. In other words, the restriction is not limited to conversations where violent intent is plain; it also covers cases where no such signs are present.
What came before
OpenAI's safety guidelines already prohibited helping users plan violent acts. The new restrictions extend that approach to teenagers and to seemingly neutral questions about firearms.
The source text was published in abbreviated form, so further details on how exactly the restrictions will be applied and when they take effect are not available here.
