The information displayed in the AIM should not be reported as representing the official views of the OECD or of its member countries.
Multiple reports highlight that AI systems, such as chatbots and generative models, have been used to obtain and organize sensitive information about dangerous pathogens and toxins, sometimes bypassing safety mechanisms. These incidents, investigated by authorities and experts, raise significant concerns about AI's role in facilitating biological weapons development and bioterrorism.[AI generated]
Why's our monitor labelling this an incident or hazard?
The article explicitly describes AI systems (chatbots and generative AI) being used in attempts to develop or plan biological weapons, including documented cases of researchers circumventing AI safeguards to obtain dangerous information. These activities represent direct or indirect involvement of AI in events that pose harm to public health and safety, fitting the definition of an AI Incident. The harms include potential injury or harm to groups of people through bioterrorism and violations of biosecurity obligations. Although some uses are legitimate research, the article focuses on the misuse and risks, with concrete examples of AI-facilitated attempts. Therefore, the event is classified as an AI Incident rather than a hazard or complementary information.[AI generated]