The UK puts about £600m into defending against AI threats

The UK Treasury announced the package on Saturday, 24 October, as part of the budget Chancellor John Healey will present next week. It adds up to about £600m. Healey called it a "distinctively British approach to AI" that does not leave workers behind.

Biosecurity gets the most. A new national programme receives £165m. Of that, £100m goes to the AI Biosecurity Lab, which will use AI to detect and track biological threats. Another £65m is for pathogen surveillance and countermeasures against the most dangerous ones. OpenAI and Anthropic will give the lab up to £200m in model credits, which means access to their AI, not cash. Google DeepMind and London-based Basecamp Research are partners.

The second part is about the security of the systems themselves. The AI Security Institute, which tests new models and writes technical standards, gets an extra £120m. £60m goes to a National Cyber Shield that will use AI agents to help defend networks, including a £25m challenge fund for UK developers. Another £70m funds a new NERV centre that will use AI to find vulnerabilities across critical infrastructure. Sky News notes that AI agents recently found long-hidden flaws in PHP, the language behind much of the web.

The rest is about home-grown tech and the regions. The Sovereign AI Fund gets £60m more to invest in UK AI companies, including robotics and manufacturing. Another £60m is for research contracts in health and AI security, £18m for a chip design academy, £15m for a prize for AI solutions, starting with health, £23m over three years for an institute studying AI's economic impact, and £30m for mayors' local projects. Separately, £10m a year goes to studying how AI will change jobs.

One caveat: model credits are not the same as money, and the sums are spread over years. How much of this is new, and how much was already planned, will only be clear from the budget.

已确认的事实

• On 24 October the UK Treasury announced a package of about £600m for AI security and development, with details due in next week's budget.^[1][2]
• £165m goes to a national biosecurity programme: £100m for the AI Biosecurity Lab and £65m for pathogen surveillance and countermeasures.^[1][2][3]
• OpenAI and Anthropic will give the lab up to £200m in model credits, with Google DeepMind and Basecamp Research as partners.^[1][3]
• The AI Security Institute gets an extra £120m.^[1][2]
• £60m goes to a National Cyber Shield, including a £25m challenge fund, and £70m to a NERV centre to find vulnerabilities.^[1]
• The Sovereign AI Fund gets £60m more, and another £60m goes to research contracts.^[1][3]
• £18m for a chip design academy, £15m for an AI prize, £23m over three years for the AI Economics Institute, £30m for mayors.^[1][2]
• £10m a year goes to studying how AI will affect jobs.^[1][2]

影响

• 已确认The UK is treating biosecurity and cyber defence as the main risks of AI.
• 我们的解读UK start-ups in AI and cyber security may win grants and public contracts.
• 我们的解读Other countries may copy the model where AI firms pay in model credits rather than cash.

仍未知

• How much of the £600m is new money and how much was already planned.
• How many years the sums cover, where no period is given.
• When the AI Biosecurity Lab will start work.

影响到谁

United Kingdom. People working in AI and cyber security, and those following the UK budget

来源

1. HM Treasury (GOV.UK): Chancellor unveils Britain's ambitious plan for AI 2026年10月24日 · 官方
2. Sky News: UK pledges to spend £600m to defend against AI health and security threats 2026年10月24日 · 媒体
3. Investing.com: UK unveils $800 million AI security funding plan 2026年10月24日 · 媒体