Artificial intelligence models capable of dramatically boosting cyberattacks on governments, businesses and critical infrastructure could emerge within months, not years, intelligence and cybersecurity agencies from the Five Eyes alliance warned in a rare joint statement on Monday.
The warning, issued by agencies from the United States, United Kingdom, Canada, Australia and New Zealand, signals growing concern among Western governments that rapidly advancing AI systems could soon transform the cyber threat landscape.
“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 alliance said in its three-page statement.
The Five Eyes – one of the world’s most influential intelligence-sharing alliances – said organisations must urgently prepare for a future in which AI accelerates the speed, scale and sophistication of cyberattacks while simultaneously offering new defensive tools.
Why the warning matters
Although the statement did not name any company or AI model, it comes amid heightened scrutiny of advanced cybersecurity-focused AI systems that can identify software vulnerabilities, generate exploit code and automate complex hacking tasks.
The warning follows recent actions by the US government to restrict access to some of the most advanced AI models developed by AI startup Anthropic. Earlier this month, Washington ordered the company to suspend access to its powerful Fable 5 and Mythos models for foreign nationals, citing national security concerns.
According to reports, Mythos is designed to identify vulnerabilities in cyber systems and has sparked concerns among policymakers that such capabilities could be misused by hostile actors to launch sophisticated attacks against governments and businesses.
The concerns are not limited to Anthropic. Intelligence officials have increasingly warned that the next generation of AI systems could significantly lower the barrier to entry for cybercriminals, allowing less-skilled actors to carry out attacks that previously required highly specialised expertise.
Cyber risk now a leadership issue?
The Five Eyes statement stressed that cybersecurity can no longer be treated as a problem confined to IT departments.
“In this environment, cyber resilience is integral to advancing business continuity, market confidence, and long-term value,” the alliance said, adding that a “whole-of-organisation and whole-of-society response” would be required.
Officials warned that cyber risk should now be viewed as a boardroom issue and a core business challenge rather than a purely technical one. They urged organisations to strengthen basic cyber hygiene measures, including rapid software patching, tighter access controls and reducing unnecessary internet exposure.
At the same time, the agencies encouraged companies and governments to use AI defensively to identify vulnerabilities faster and respond more effectively to cyber incidents.
A narrowing window
Cybersecurity experts say the warning reflects a broader concern that AI development is moving faster than many organisations can adapt.
The Five Eyes agencies noted that assumptions about cyber risk may become outdated within months as AI capabilities continue to improve rapidly. They warned that organisations delaying investments in cyber resilience could find themselves increasingly vulnerable as more powerful AI tools become widely available.
