The information displayed in the AIM should not be reported as representing the official views of the OECD or of its member countries.
The US government ordered Anthropic to suspend access to its advanced AI models, Fable 5 and Mythos 5, for all foreign nationals due to a discovered jailbreak vulnerability posing national security risks. Anthropic complied but disputes the severity, calling the directive a misunderstanding and seeking to restore access.[AI generated]
Why's our monitor labelling this an incident or hazard?
The article explicitly involves AI systems (Claude Fable 5 and Mythos 5) and discusses their suspension due to security concerns raised by US authorities. The concerns relate to vulnerabilities that could be exploited to bypass safeguards, potentially leading to cyberattacks or unauthorized access to sensitive information, which would constitute harm. Since no actual harm or incident has occurred yet, but the risk is credible and serious enough to warrant suspension, this fits the definition of an AI Hazard. The event does not describe realized harm or violations, so it is not an AI Incident. It is not merely complementary information or unrelated news, as the focus is on the plausible risk of harm from the AI systems' vulnerabilities.[AI generated]