Five Eyes warns advanced AI could overwhelm cyber defenses in months

Five Eyes cyber agencies warned on June 22 that frontier AI models could transform offensive and defensive cyber capabilities in months, not years, in a rare joint statement from CISA, the UK National Cyber Security Centre, the Canadian Centre for Cyber Security, the Australian Cyber Security Centre and New Zealand’s Cyber Security Directorate. The agencies said risk assumptions can go stale quickly and pressed leaders to act now.

The statement urged organizations to tighten foundational security, empower cyber leaders and treat cyber resilience as a business responsibility rather than a narrow technical task. It also said companies should use AI deliberately to strengthen defenses, not just squeeze more efficiency out of existing systems. The message was blunt: AI is no longer a future issue to be studied later, but a near-term force that can change how fast attackers move and how quickly defenders fall behind.

Officials said the threat is already visible. The UK National Cyber Security Centre has said the most advanced tools already make it easier, faster and cheaper for attackers to find and exploit weaknesses, including zero-day vulnerabilities. The Canadian Centre for Cyber Security said the warning was issued now because it is seeing real, recent shifts in how AI tools are being used to speed the discovery and exploitation of vulnerabilities. Chris Krebs, the former CISA director, called the warning “pretty alarming.”

The joint statement offered few new technical prescriptions and largely repeated standard security guidance, including patching software quickly and avoiding unnecessary exposure of systems to the internet. That restraint underscored the bigger problem facing governments and companies: the core defenses many organizations still rely on were built for a slower threat environment, while AI is accelerating both attacker tradecraft and the scale at which it can be deployed.

The warning also landed against a wider policy push. CISA and international partners released a guide on the secure adoption of agentic AI services on May 1, and the White House issued an executive order on June 2 focused on advancing AI innovation and security. Together, the moves show governments trying to get ahead of a shift in which organizations that fail to adapt could face operational, financial and strategic disadvantage as AI reshapes the cyber battlefield.