Five Eyes Warn Frontier AI Could Upend Cybersecurity Within Months

Five Eyes intelligence alliance says frontier AI could trigger a cyber paradigm shift within months

"Cyber risk is a core business risk and a leadership responsibility"

Reuters says warning signals mounting concern among intelligence agencies over Anthropic and OpenAI models

The Five Eyes intelligence alliance of the US, UK, Canada, Australia and New Zealand warned that AI-driven cybersecurity threats could become reality within months and called for immediate action.

In a joint statement released on July 22, the alliance said cutting-edge frontier AI models could fundamentally reshape cyber capabilities for both attack and defense beyond current industry expectations.

"That paradigm shift will happen in just months, not years," the statement said.

The Five Eyes urged national leaders to act immediately. It called for assessments of AI-related risks, preparedness and accountability; prioritizing basic cybersecurity practices and control systems; giving cybersecurity leaders full authority and resources; and staying actively engaged with evolving threat indicators and guidance.

The intelligence agencies added that the response must extend across organizations and society as a whole.

"Cyber risk is no longer just a technical issue. It is a core business risk and a leadership responsibility," they said.

The statement did not provide specific examples or evidence of AI-based cyberattacks.

The market sees the warning as an extension of broad security concerns that have persisted since the so-called Mytos shock, which raised fears that AI could fundamentally disrupt the existing security ecosystem.

Still, the statement carries added weight because intelligence agencies directly warned about cybersecurity threats tied to AI.

Reuters said the warning shows global intelligence agencies' concerns over advanced AI models such as Anthropic's Mytos and OpenAI's GPT 5.5-Cyber have reached a critical level.

The statement did not name any specific model, but it is widely viewed as aimed at Anthropic's Mytos, which has recently been subject to export controls.

On July 12, Anthropic said the US government had issued export-control guidance fully blocking foreign nationals from accessing Claude Mytos 5 and Claude Fable 5.

South Korea's government and some companies recently joined Project Glasswing, which provides access to Mytos, but the latest move by the US administration has constrained use of the model.

Park Sang-kyung, Hankyung.com reporter highseoul@hankyung.com