The UK government has announced more than £600 million in artificial intelligence investment covering biological threat detection, automated cyber defence, safety testing and support for domestic technology companies. The package, unveiled on October 24 ahead of the Budget, combines programmes to address risks from powerful AI systems with funding to use the technology for national defence.
The measures include a £165 million National AI Biosecurity Initiative, an additional £120 million for the AI Security Institute and new programmes to find software vulnerabilities and deploy AI agents in cybersecurity. Separate allocations will support British AI businesses and research procurement, reflecting the government’s aim to reduce dependence on overseas technology.
Within the biosecurity initiative, £100 million will establish an AI Biosecurity Lab. The government says the lab will use advanced models to detect biological threats, strengthen surveillance and support the development of medical countermeasures. The remaining £65 million will fund biosurveillance and work on treatments or other countermeasures for priority pathogens.
The programme addresses the dual-use nature of AI in biology. Models that help researchers investigate disease and develop medicines may also make some dangerous research easier to undertake. The lab is intended to examine both the defensive applications of those capabilities and the risks associated with their misuse.
OpenAI and Anthropic have committed up to £200 million in model access and computing credits to support the programme. That commitment is distinct from government funding. Google DeepMind and British life-sciences company Basecamp Research will also take part in research collaborations.
The additional funding for the AI Security Institute will expand its evaluation work and support the development of technical standards for advanced systems. The institute examines model capabilities and weaknesses in safeguards, including through cybersecurity assessments, tests of autonomous agents and research into potentially dangerous uses.
A further £60 million will go towards a National Cyber Shield, which will use agentic AI for defensive tasks. Unlike chatbots that mainly respond to individual prompts, agents can perform sequences of actions with less direct human supervision. In cybersecurity, those capabilities can support investigations, testing and responses to suspicious activity, although their permissions and actions still require careful control.
The package also includes a £25 million challenge fund for British AI and cybersecurity companies. The government is considering a regulatory sandbox where agentic security tools could be tested under controlled conditions across critical infrastructure, government and industry. That testing arrangement remains a proposal rather than an established programme.
Another £70 million will establish the National Exposure and Real-Time Vulnerability Centre, known as NERV. It will use AI to identify cybersecurity weaknesses across critical infrastructure and the wider economy before attackers exploit them. The government says existing AI-assisted pilots have already found and addressed more than 1,000 vulnerabilities.
Finding flaws is only one part of vulnerability management. Organisations must also assess which findings represent genuine risks, prioritise repairs and deploy fixes without disrupting essential services. Faster discovery can increase the volume of work facing security teams unless assessment and remediation keep pace.
Beyond the defensive programmes, the government will provide another £60 million to the Sovereign AI Fund for direct investment in British businesses, including companies developing robotics, autonomous systems and industrial AI. A separate £60 million research and development procurement programme will focus on healthcare, AI security and safe adoption.
Sources: The420.in