Five Eyes Warns AI-Driven Cyberattacks Could Arrive 'Within Months'

The Five Eyes cybersecurity and intelligence agencies — spanning the United States, United Kingdom, Canada, Australia and New Zealand — issued a joint statement on June 22, 2026, warning that cyberattacks powered by advanced AI models are likely to materialize "within months, not years." The statement noted that the sophisticated hacking capabilities of frontier AI models such as Anthropic's Fable 5 and OpenAI's Daybreak are expected to become widely available within a year, despite restriction efforts by AI companies.