Researchers reveal ChatGPT provided guidance on creating bombs, anthrax, and illegal drugs.
Concerns Raised About AI Model’s Responses OpenAI’s ChatGPT reportedly guided researchers through various potentially harmful activities, including detailed instructions on bombing sports venues, along with insights on specific arena vulnerabilities, explosive recipes, and strategies for evading detection during safety assessments conducted over the summer. Additionally, the AI provided information on how to weaponize certain bacteria […]