Five Eyes intelligence agencies warn: destructive AI could emerge in months, not years
The intelligence and cybersecurity agencies of Australia, the US, UK, New Zealand, and Canada issued a rare joint statement warning that powerful artificial intelligence models capable of devastating cyberattacks against governments and businesses could appear within months.
In a rare joint statement, the Five Eyes intelligence and cybersecurity agencies urged leaders and businesses to act now. They noted that while AI will eventually help improve cyber defense, it simultaneously increases the speed, scale, and sophistication of cyber threats.
The document emphasizes that advanced AI models will surpass current industry expectations and fundamentally change both offensive and defensive capabilities in cyberspace. This is a matter of months, not years. The agencies highlighted that breakthroughs in AI model development show the technology will lower barriers for attackers and increase the speed and complexity of attacks.
The statement calls for a comprehensive response at both the organizational and societal levels. Although no specific AI models or companies are named, attention is focused on Anthropic's advanced tools. Olivia Shen, an expert from the US Studies Centre at the University of Sydney, warned that next versions of the Mythos or Fable level could appear soon, and that there may be equally advanced models being developed by China or other actors.
Background: In May 2026, the Five Eyes published joint recommendations describing 23 categories of risks related to autonomous AI systems. In early June 2026, Anthropic released its newest models, Fable and Mythos, with the latter having enhanced capabilities to find vulnerabilities in code. According to The Economist, on June 11, US Senate Intelligence Committee Vice Chair Mark Warner stated that Mythos hacked nearly all secret systems of the NSA and US Cyber Command within hours. On June 12, the US Commerce Department issued an emergency export control directive, requiring Anthropic to block access to Fable and Mythos for foreign nationals.
