AI models that can take down governments and business months away, rare Five Eyes statement warns

Cyber intelligence agencies for the Five Eyes alliance have issued a rare joint warning that frontier AI models capable of destabilizing governments and businesses are only months away, urging leaders to "act now" to bolster cyber resilience [1]. The statement, released late Monday Sydney time by signals agencies from Australia, the United States, the United Kingdom, New Zealand, and Canada, marks an unusual public intervention [1]. The agencies assessed that while artificial intelligence—the capability of computational systems to perform tasks associated with human intelligence [2]—will improve cyber defense over time, it simultaneously accelerates the speed, scale, and sophistication of threats [1]. "Frontier AI models are anticipated to exceed current industry expectations, fundamentally transforming both offensive and defensive cyber capabilities. The timeline is not years, it is months," the warning stated [1]. The alert comes weeks after the Trump administration blocked "foreign nationals" from using Anthropic's advanced AI model, Fable, citing national security advice [1]. Anthropic, which employs AI alignment researcher Jan Leike who joined the firm from OpenAI in May 2024 [8], has also developed a more powerful model called Mythos. Claude Mythos is a large language model designed to find vulnerabilities in software, but Anthropic has not officially released it to the public due to safety and misuse concerns [6]. Both Fable and Mythos were suspended for use by foreign nationals in June [1]. The Five Eyes agencies warned that the leaps in AI models will lower barriers for malicious actors and increase the complexity of attacks [1]. "A whole-of-organisation and whole-of-society response is required," the statement continued. "Cyber risk can no longer be treated as a purely technical issue. This is a core business risk and leadership responsibility" [1]. Olivia Shen, a national security and AI expert at the University of Sydney's United States Studies Centre, cautioned that the focus on Anthropic may obscure a broader threat. "I think we have to anticipate that the next Mythos or the next Fable is just around the corner," Shen said [1]. She noted that other state and non-state actors, including China, could be developing equally advanced models outside of public view [1]. The Albanese government signed a non-binding memorandum of understanding with Anthropic in March as part of its national AI plan, which promotes a light-touch regulatory approach to capture economic benefits from the technology [1].