Five Eyes agencies warned that frontier AI could reshape cyberattacks within months, pushing governments and companies to treat resilience as a leadership risk.
Key Points:
- Five Eyes agencies said frontier AI could alter cyber offense and defense within months.
- The warning followed U.S. limits on foreign access to Anthropic models Fable and Mythos.
- Newer AI systems can find and exploit vulnerabilities faster.
Five Eyes Warning
Signals intelligence agencies from Australia, the United States, Britain, New Zealand and Canada issued the rare public statement late Monday in Sydney, urging leaders to act now.
The warning followed the Donald Trump administration’s June decision to block foreign nationals from using Anthropic’s Fable model, while also suspending access to Mythos.
AI will improve cyber defense over time, but it also raises the speed, scale and sophistication of attacks, the agencies said.
“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 statement said.
Cyber risk can no longer sit only with technical teams.
Also Read: Ripple Wins Preliminary MiCA Approval As EU Crypto Race Tightens
Anthropic AI Risk
Generative AI can find weaknesses in cybersecurity systems, help exploit them and help fix them, making the same tool useful to defenders and attackers.
Olivia Shen, a national security and AI expert at the University of Sydney’s United States Studies Centre, said the newest systems can generate exploits more effectively.
“What’s different about the latest [AI models] ones is they’re very good at generating exploits,” Shen said.
Anthropic’s Fable 5 has drawn attention as a more public-facing version of Mythos, an advanced system released earlier this year and limited to vetted users because of misuse concerns.
“I think we have to anticipate that the next Mythos or the next Fable is just around the corner,” Shen said, adding that other actors may be developing similar systems.
In March, the Albanese government signed Anthropic as the first company in its national AI plan, a non-binding arrangement focused on progress sharing and safety.
The Five Eyes alliance dates to the period after the second world war, and its joint public warnings are uncommon, making this statement a sign that AI cyber risk has moved beyond routine technical guidance.
Read Next: OpenAI’s $100B ChatGPT Ads Plan Meets A Cold Cannes Lions Reality