Intel agencies: Frontier AI models will reshape cybersecurity faster than expected

The Five Eyes intelligence agencies warn that advanced "frontier" AI models—already appearing in commercial, open-source, and illicit channels—are months away from broad public availability and could rapidly transform offensive and defensive cyber capabilities; the statement emphasizes existing systemic weaknesses (legacy systems, slow patching, unnecessary connectivity, and poor access controls) and urges immediate action on core cybersecurity fundamentals and defensive AI programs.