Safeguarding our national security is the first duty of any government, according to the Chancellor of the Exchequer John Healey in the Budget 2026. He announced an additional £120m for the AI Security Institute, and £165m for a new National AI Biosecurity Initiative.
On AI, the Budget stated: “AI offers profound opportunities, but these can only be realised if it is trusted and secure. The risks increase as AI becomes more capable and widely deployed.” A new AI Cyber Security Centre of Excellence will be set up in Manchester. A £70m National Exposure and Real-Time Vulnerability (NERV) Centre at the UK official National Cyber Security Centre (NCSC) to aid the UK’s ability to understand and reduce national cyber exposure.
The Budget included an investing of £154m to strengthen cyber defences and £40m to establish a National Agentic Security Unit to investigate and respond to rogue agentic incidents. As for artificial intelligence (AI), defined as ‘an age-defining general purpose technology’, the UK Government will invest an additional £140m in the Sovereign AI Unit to back promising UK AI companies to scale and succeed here. It will also set out a new Data Centre Covenant, establishing the tests that strategic data centres are expected to meet, including the contribution they make to communities and the UK’s economic interests.
Cyber shield
Some £60m will go into the first phase of a national Cyber Shield to develop agentic cyber defence capabilities, partnering with the UK cyber security industry. This will include a £25m AI and Cyber Security Challenge Fund to support UK innovation, alongside scoping for a regulatory deployment sandbox to help deployment of new technologies.
Defence
Among announcements in defence, the Budget gave £400m to equip the Army with the latest drone tech.
Liam Hutcheson, Country Director, UK and Canada, for MyDefence welcomed this spend, but questioned how quickly will that funding translate into the advanced intelligence our Armed Forces need now. He said: “While investment in equipment and infrastructure remains essential, the nature of modern warfare is changing faster than traditional defence programmes can respond. Artificial intelligence is transforming the battlefield, where adversaries are continually adapting how they communicate, operate and exploit the electromagnetic spectrum.
“The challenge is no longer simply detecting a drone. It is understanding the increasingly complex electromagnetic environment in which threats operate, including signals that existing detection systems may not recognise. This demands a shift in defence investment priorities. The Armed Forces need AI-powered capabilities that can analyse unfamiliar signals in real time, operate at the tactical edge and turn electromagnetic intelligence into actionable insights for front line personnel. The UK has recognised the importance of AI, but ambition alone is not enough. Investment must translate into capabilities that frontline personnel can deploy to understand and respond to emerging threats in real time. Without this, defence readiness risks falling behind the pace of technological change.”
Comment
James Mackay, CEO at MetaCompliance, said: “The bulk of the government’s security funding, including the £60m National Cyber Shield, is focused at national-level cyber defence, and it’s a welcome start. We don’t know yet when these tools will be deployed or how far they’ll act without human oversight, but one thing is clear, training the people who use them will be key to their success. On the ground, the recent PHP flaws have left IT leaders under mounting pressure to patch and protect their own systems.
“Using AI to patch vulnerabilities is a positive step, but it only solves part of the problem. Many of today’s cyber attacks start with human error, and AI is making attacks aimed at people much harder to spot. CISOs are navigating a fast-moving threat landscape, and often the toughest part is preparing their people. As the government’s package speeds up AI adoption, safeguards and policy will take time to catch up. In the meantime, some employees are turning to unapproved tools. The extra £120m for the AI Security Institute will help shape what safe adoption looks like. Employers can help close that gap too, by showing staff how to use AI safely at work and how to spot and report AI-driven attacks linked to the PHP threat and other major hacks, like fake ‘urgent patch alerts’ posing as IT support.
“Organisations can start with up-to-date AI policies, clear guidance employees can follow under pressure and training tailored to each person’s role and risk profile, with its impact measured. Security training often falls flat when it feels generic. That’s why it’s vital to measure behaviour change, track what’s working across every part of the business, and share behavioural risk with the board.”
