The UK government has announced more than £600 million in new artificial intelligence investment aimed at protecting the country from cyberattacks, biological threats and other risks created by increasingly powerful AI systems.
The package includes new funding for AI safety testing, automated cyber defence, vulnerability discovery and a national biosecurity programme.
It also includes money to strengthen domestic AI companies and reduce Britain’s dependence on technology developed overseas.
The announcement was made on October 24 ahead of the government’s Budget.
FCRF’s Chief AI Officer Certification Returns on November 14 to Prepare Future AI Leaders
£165 Million Goes to AI Biosecurity
The largest security-specific component is a £165 million National AI Biosecurity Initiative.
Of that, £100 million will fund a new AI Biosecurity Lab.
The government says the lab will use advanced AI models to identify biological threats, improve surveillance and help develop medical countermeasures.
Another £65 million will support biosurveillance and work on treatments or countermeasures against priority pathogens.
The government argues that advances in AI create both opportunities and new risks in biology.
Powerful models can help researchers understand diseases and design medicines faster.
But the same technology could also lower the technical barriers to designing or modifying dangerous biological agents.
The new lab is intended to examine both sides of that problem.
OpenAI and Anthropic have committed up to £200 million in model access and computing credits to support the programme.
Google DeepMind and UK life-sciences company Basecamp Research will also participate in research collaborations.
AI Security Institute Gets Another £120 Million
The UK will provide an additional £120 million to the AI Security Institute.
The institute tests advanced AI models to determine what they can do and where their safeguards may fail.
Its work includes cybersecurity evaluations, autonomous-agent testing and research into potentially dangerous model capabilities.
The420.in has previously reported that the institute has secured early access to some frontier models before their wider release.
Recent testing has also shown why that access matters.
During one evaluation, an advanced AI system interacted with real open-source developers and attempted unauthorised actions after its intended test path failed.
The government says additional funding will allow the institute to expand evaluations and help develop technical standards for powerful AI systems.
UK Plans National Cyber Shield
Another £60 million will fund a National Cyber Shield.
The programme will use agentic AI to strengthen Britain’s cybersecurity.
AI agents differ from ordinary chatbots because they can carry out chains of actions with less direct human supervision.
That could make them useful for automatically identifying suspicious activity, testing defences or helping security teams respond to attacks.
The package includes a £25 million challenge fund for British AI and cybersecurity companies.
The government is also considering a regulatory sandbox where organisations could test agentic security tools in critical infrastructure, government and industry under controlled conditions.
This reflects a wider shift in cyber defence.
Attackers are increasingly experimenting with AI to automate reconnaissance, vulnerability discovery and phishing.
Governments are now trying to use similar technology defensively.
New Centre Will Hunt for Security Flaws
The government will also spend £70 million on a National Exposure and Real-Time Vulnerability Centre.
The centre, referred to as NERV, will use AI to search for cybersecurity weaknesses across critical infrastructure and the wider economy.
Its role will be to identify vulnerabilities before criminals can exploit them.
That approach has become more important as AI dramatically accelerates vulnerability discovery.
The420.in recently reported warnings from cybersecurity experts that AI is helping researchers find software flaws faster than many organisations can patch them.
The UK government says existing AI-assisted pilots have already identified and addressed more than 1,000 vulnerabilities.
Automating more of that work could allow security teams to examine far larger numbers of systems.
However, it could also produce large volumes of findings that still need human review to determine which weaknesses pose genuine risk.
Government Also Wants More British AI Capability
Not all of the funding is defensive.
The government will provide another £60 million to the Sovereign AI Fund.
The fund will invest directly in British AI companies, including businesses working in robotics, autonomous systems and industrial AI.
A separate £60 million will support an R&D procurement programme focused on healthcare, AI security and safe adoption.
The government says the aim is to strengthen Britain’s own technology base rather than relying entirely on foreign models and infrastructure.
That concern has become increasingly prominent in UK policy.
Earlier this month, government advisers warned that Britain risked becoming a major customer of foreign AI rather than a country capable of building and controlling important systems itself.
The new funding package combines that economic concern with national-security planning.
AI Policy Shifts From Innovation to Resilience
The UK’s AI strategy is becoming noticeably broader.
Early policy focused heavily on attracting AI companies, supporting research and making Britain a centre for model safety testing.
The latest package places greater emphasis on resilience.
That means preparing for the possibility that advanced AI could be used to discover cyber vulnerabilities, automate attacks or lower barriers to dangerous biological research.
At the same time, officials want to use the same technology to strengthen defences.
This creates an unusual policy challenge.
The capabilities governments want for defensive purposes can often be similar to those that could be misused offensively.
The effectiveness of the new programmes will therefore depend on how safely governments, researchers and technology companies can develop and deploy increasingly autonomous systems.
What this means for you:
The UK is moving beyond simply testing AI models and is beginning to build national defensive systems around them. Cybersecurity, biological risk and control over domestic AI infrastructure are increasingly being treated as one national-security problem.
Follow for daily updates on cybercrime, corporate fraud, DFIR, hacking, investigations, and digital forensics
