Hundreds of ChatGPT Users Asked About Bioweapons and Poisons — It Answered

A report has revealed that hundreds of ChatGPT users asked the AI model questions about bioweapons and poisons, and received answers. Experts note that for now, reliable books remain a more dangerous source of such information than current AI models, but the technology is advancing rapidly. The findings reignite debate about safety guardrails in large language model systems.