Five Eyes Warns of AI Models Capable of Government, Business Takedowns Within Months

The statement highlighted that the timeline for such capabilities is measured in months, not years, and that cyber resilience is crucial for business continuity, market confidence, and long-term value. The agencies emphasized that cyber risk should no longer be treated as a purely technical issue but as a core business risk and leadership responsibility.

This intervention follows the US government's decision earlier this month to block foreign nationals from using Anthropic's highly anticipated AI model, Fable. Anthropic's models, including Fable 5 and Mythos, have raised concerns due to their advanced capabilities in detecting cyber vulnerabilities, leading to their suspension for use by foreign nationals in June. Mythos, in particular, is only available to vetted organizations due to fears of exploitation.

Experts like Olivia Shen from the University of Sydney's United States Studies Centre believe that the development of advanced AI models is ongoing globally, with potential for similar or even more advanced systems emerging from countries like China, as well as other states and companies. The Australian government has engaged with Anthropic, signing a memorandum of understanding as part of its national AI plan to promote safety and share details of AI progress, adopting a light-touch regulatory approach.