Britain should prepare for an AI-powered cyberattack on its critical national infrastructure within two years, the National Cyber Security Centre said on Tuesday. In its annual report, the agency said it was highly likely that a ransomware attack using AI-powered malware would hit critical infrastructure by 2028. The center’s chief executive, Richard Horne, said China was likely the best equipped to carry out such operations.
What Horne warned
Horne said organizations and supply chains had got away with doing technology on the cheap for decades, leaving them exposed. “In the coming months, we face a real prospect of an unprecedented rise in the number of cyber incidents,” he was expected to say in a speech on Tuesday, warning that vulnerabilities tolerated can now be exploited at unimaginable scale.
Horne was due to say Russia was making sustained attempts to penetrate UK military networks, while Iran was carrying out malicious cyber operations globally. Moscow, Beijing and Tehran have repeatedly denied such accusations, and the report’s claims about China rest on the agency’s assessment, not on published evidence.
Our read: the warning is about scale, not magic
Horne’s most important distinction is easy to miss: he is not saying AI has invented new ways to attack. His argument is that weaknesses organizations tolerated for years can now be discovered and exploited far faster, and across far more targets at once. That reframes the danger from a futuristic weapon to a present-day backlog of unfixed flaws.
The warning also lands as AI moves deeper into government systems, where oversight questions follow it. A personal Claude agent reportedly ran for two months on a Pentagon network, showing how fast experimental tools can reach sensitive work before the rules around them catch up.
The report’s warning is framed as preparation, not a dated prediction of a specific attack: Britain should be ready within two years, and a ransomware attack using AI-powered malware is highly likely by 2028. That leaves room for prevention — the very action Horne is demanding — rather than presenting the outcome as fixed.
Why It Matters
Critical infrastructure is the highest-stakes target in cyber conflict because the damage is physical: power, water, food and hospitals depend on the same networks attackers probe. Horne’s warning to act now, or explain later why food was not on shelves and lights were not on, is aimed squarely at boards that still treat security as an IT cost.
That scale is not hypothetical. CrowdStrike has described a China-linked hacking group using an AI agent framework to run intrusions against more than 200 organizations in at least 14 countries, with post-exploitation activity confirmed at nine. The campaign shows agents moving from assisting one intrusion to carrying out many in parallel.
The timing matters too. The warning follows hacking attacks on supermarkets across Europe that exposed weaknesses in IT systems, a concrete reminder that disruption does not need a nation-state cyber weapon to empty shelves. AI simply lowers the cost of finding the next weak link, and raises the speed at which it can be pulled.
Background
The National Cyber Security Centre is part of GCHQ, Britain’s signals intelligence agency, and its annual report is one of the few regular public assessments of the country’s cyber threat picture. Its warnings carry weight precisely because the agency rarely predicts specific attacks, preferring to describe trends and urge preparation before incidents arrive.
Source: Reuters report
0 Comments