On November 3, 2026, the UK National Cyber Security Centre, part of GCHQ, published its annual State of the Cyber Nation review, covering September 1, 2025 to August 31, 2026, presented in a speech by chief executive Richard Horne. Its headline judgement is that AI-powered malware is “highly likely” to be used in a ransomware attack on UK critical national infrastructure by 2028.
The NCSC also judges that within two years hostile actors’ use of AI is “highly likely” to increase the scale and efficiency of ransomware, social engineering and vulnerability exploitation, and it expects an “unprecedented rise in the number of cyber incidents”. Its framing is that AI is not inventing new attacks but letting attackers find and exploit long-tolerated weaknesses at scale, after decades of what it calls “tech on the cheap”.
The release links the review to investments announced the previous week: 60 million pounds for the first phase of a national Cyber Shield, 70 million pounds for a National Exposure and Real-Time Vulnerability (NERV) Centre, and a new AI Cyber Security Centre of Excellence run by the NCSC in Manchester. Its recommendations include urgent investment, not connecting technology to the internet unnecessarily, and treating Cyber Essentials as “the starting point for cyber security, not the end goal”.
Why it matters: it puts a dated, official probability judgement from a national cyber agency on AI-enabled attacks against critical infrastructure, which turns a frequently cited risk into a planning assumption for UK operators.
What it does not show: the release gives no incident counts or evidence for the 2028 estimate, and “highly likely” is an intelligence assessment, not a measurement. It does not attribute any past attack on UK infrastructure to AI-powered malware.