The National Cyber Security Centre, operating under GCHQ, has warned that artificial-intelligence-enhanced malware is highly likely to be deployed in ransomware assaults on the United Kingdom’s critical national infrastructure before 2028. The warning covers facilities such as power stations, airports, hospitals and other services spanning energy, transport, defence, food, health and finance. The agency’s annual review stresses that the threat is imminent and that AI could change the scale of attacks.
Ransomware remains a core weapon for criminal groups, many linked to former Soviet states, that first breach networks, encrypt data and demand cryptocurrency for decryption. Recent UK victims have included the national postal service, an outsourcing firm and several local authorities. The NCSC says AI will streamline the preparatory phases of such campaigns, automating target scouting, vulnerability identification and even code generation, thereby lowering the barrier for sophisticated attacks.
The government’s Be Ready campaign has urged households to store up to nine litres of bottled water and three days’ worth of food for emergencies, a measure announced by Defence Secretary Wes Streeting as preparation for possible state-sponsored terrorism. The advisory reflects growing concern that hostile powers could exploit AI-driven cyber tools against everyday infrastructure.
NCSC chief executive Richard Horne is set to address the public in a speech accompanying the review, warning that AI will unleash a surge of cyber incidents. He will state, “In the coming months, we face a real prospect of an unprecedented rise in the number of cyber incidents.” Horne will link the danger to AI’s ability to expose long-ignored weaknesses in IT systems.
An AI-assisted discovery last month exposed a critical flaw in PHP, a programming language powering roughly 70 % of global websites. Surrey University cybersecurity professor Alan Woodward said the NCSC alert highlights the need for nations and enterprises to obtain cutting-edge AI tools, noting, “AI can help people find these flaws and patch them, but if we don’t get the access the bad guys will beat them to it.” Leading model providers such as OpenAI and Anthropic have already granted limited advanced access to selected firms for vulnerability hunting.
Horne will also note that British military communications networks endure daily, persistent attacks, with Russia identified as a frequent actor that deliberately stays below the threshold of armed conflict. He will cite recent incidents in Europe where hackers raised supermarket refrigerator temperatures above 50 °C, using the episode as a warning that organisations must invest heavily in cyber resilience to counter AI-enhanced threats.