The information displayed in the AIM should not be reported as representing the official views of the OECD or of its member countries.
Japan is negotiating with Anthropic for access to its advanced AI model, Claude Mythos, amid concerns it could be exploited for cyberattacks on critical infrastructure. The government has ordered a comprehensive cybersecurity strategy review to address the risks posed by the model's vulnerability-finding capabilities and potential misuse.[AI generated]
Why's our monitor labelling this an incident or hazard?
The article involves an AI system (Claude Mythos) capable of identifying software vulnerabilities, which could plausibly be exploited for cyberattacks causing harm to critical infrastructure or systems. However, no realized harm or incident is described; the concerns are about potential misuse and future risks. The government's negotiation to access the AI model for defensive purposes and the establishment of countermeasures indicate a response to a credible hazard. Therefore, this event fits the definition of an AI Hazard, as it plausibly could lead to an AI Incident but no incident has yet occurred.[AI generated]