The information displayed in the AIM should not be reported as representing the official views of the OECD or of its member countries.
The United States and Israel used advanced AI systems, including Project Maven, to rapidly identify and attack over a thousand targets in Iran, resulting in civilian casualties and the death of Iran's supreme leader. Reports highlight that algorithmic errors in AI-driven targeting accelerated attacks and contributed to wrongful strikes on civilian sites.[AI generated]
Why's our monitor labelling this an incident or hazard?
The article explicitly describes the use of AI systems in military targeting and attack execution, which directly led to harm including civilian deaths and destruction. The involvement of AI in accelerating decision-making and target selection is clear, and the reported errors in AI algorithms plausibly caused wrongful attacks on civilian sites. These outcomes constitute injury and harm to people and potential violations of human rights, fulfilling the criteria for an AI Incident. The harm is realized, not just potential, and the AI systems' role is pivotal in the chain of events leading to these harms.[AI generated]