THE CRUNCH
Britain's National Cyber Security Centre has warned that artificial intelligence is likely to give attackers an advantage over defenders. Dave Chismon, the NCSC’s chief technology officer, said in a blog post that automated defences cannot yet match the freedom attackers have to use AI tools. He argued that offensive work offers clear success signals, such as an exploit working or malware communicating, while defence
offers no such clear signal and carries a high risk of causing disruption if a patch or rule fails. Chismon cited security researcher Halvar Flake to explain that offensive problems are technical, whereas defensive problems are political, meaning humans must weigh each action and answer for it. The NCSC is developing a capability called Cyber Shield to deploy agentic AI systems across government networks, but Chismon cautioned that autonomous defensive actions are not yet ready to be relied on. He suggested organisations start with low-risk uses, such as having AI summarise threat intelligence, while keeping traditional security improvements in place.
The warning comes amid broader concerns from the Five Eyes intelligence alliance that frontier AI could transform cyber operations within months rather than years. While software makers have issued patches at record rates since the summer, a comparable rise in cyberattacks has not yet been observed. Chismon acknowledged that making autonomous defensive actions safe enough to deploy remains an unsolved problem and will require time and research. The NCSC will soon publish an 'AI for Cyber Defence' research agenda to help address these challenges.