Cutting-edge artificial intelligence technology is poised to supercharge offensive hacking capabilities, and urgent action is needed to face up to the threat, US, UK, Canadian, Australian and New Zealand officials have said.
“Frontier AI models are anticipated to exceed current industry expectations, fundamentally transforming both offensive and defensive cyber capabilities,” the intelligence alliance commonly known as the Five Eyes said in a three-page statement on Monday.
Washington — US cybersecurity authorities issued a fresh warning about the rising risk of AI-powered cyberattacks, while offering limited new guidance to defenders.The joint statement, signed by several federal agencies including the Cybersecurity and Infrastructure Security Agency (CISA), largely repeated standard cybersecurity recommendations. Officials advised organizations to promptly patch vulnerable software and avoid exposing systems to the internet unless absolutely necessary.
The advisory also encouraged the use of artificial intelligence to strengthen defenses — particularly for identifying weaknesses faster and responding more rapidly to security incidents.
The warning reflects growing official concern over advanced AI models, such as Anthropic’s Mythos and OpenAI’s GPT-5.5-Cyber.
These systems are reportedly capable of enabling users to carry out complex and potentially highly damaging hacks in a short time.Earlier this month, Anthropic was required to disable a version of its Mythos model after the US government suspended access for foreign nationals over national security concerns.
Around the same period, CISA — one of the agencies behind Monday’s statement — shortened the deadline for federal agencies to address critical vulnerabilities in their networks from longer periods to just three days, explicitly citing the heightened threat posed by AI.
The statement comes amid broader worries that rapid advances in AI could significantly lower the barrier for both state and non-state actors to conduct sophisticated cyber operations.