The information displayed in the AIM should not be reported as representing the official views of the OECD or of its member countries.
Anthropic revealed and prevented several attempts to misuse its Claude AI models for developing biological weapons, autonomous military drone swarms, cyber espionage, and disinformation campaigns. The incidents involved actors from China, Russia, Iran, and Yemen, highlighting significant risks from AI-enabled malicious activities.[AI generated]
Why's our monitor labelling this an incident or hazard?
The article explicitly describes the use and attempted misuse of AI systems (Anthropic's Claude models) in ways that have directly led to or could lead to significant harms: development of biological weapons (harm to health), autonomous military drone swarms (potential harm to people and infrastructure), and disinformation campaigns (harm to communities). The company's intervention to prevent these harms confirms the AI system's involvement in incidents. The presence of realized or ongoing harms and the direct link to AI system use classify this as an AI Incident rather than a hazard or complementary information.[AI generated]