🔍 Read the full analysis: Britain Pays More To Test Powerful AI, But Not Ten Times More on ThorstenMeyerAI.com
Get business pricing on tech for your team
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
TL;DR
The UK Budget confirmed an additional £120 million for the AI Security Institute across 2026-27 to 2029-30, averaging about £30 million a year. Using the source’s figures, UK annual funding is about £96 million against roughly $15 million for the US testing centre, an estimated eight-to-ninefold gap at current exchange rates—not more than tenfold; the comparison changes against a larger US budget request.
The UK Budget confirmed £120 million in additional funding for the AI Security Institute (AISI) over four financial years, from 2026-27 to 2029-30. The commitment strengthens Britain’s capacity to evaluate advanced AI systems before release, but the annualised comparison with the US testing centre does not support the claim that the UK top-up alone is more than ten times its American counterpart’s annual budget.
The new allocation averages about £30 million a year, based on dividing the total across the four-year period. The Budget does not specify how the money will be distributed year by year. AISI says its existing funding is £66 million per financial year. Combining that figure with the average annual top-up gives an estimated £96 million a year for the UK institute, although that calculation is not a published annual budget schedule.
The source estimates the UK total at about $128 million annually, compared with roughly $15 million for the US centre. On that basis, Britain’s annual funding is about eight or nine times the US figure, depending on exchange rates and rounding. The comparison is different against the White House’s requested $27 million for the US centre in fiscal 2027: the UK estimate would be less than five times as large. That US request is not yet an enacted appropriation.
The funding is part of a wider UK package that the Treasury puts at £700 million. It also includes £165 million for a National AI Biosecurity Initiative and £40 million for a National Agentic Security Unit, intended to detect and respond to rogue agentic systems. AISI, established in 2023 and part of the Department for Science, Innovation and Technology, says it has more than 100 technical staff and priority access to over £1.5 billion in public computing resources.
Infographic 01 · The money, made comparable
Britain spends about 8 to 9 times more per year. Not “10 times” from the top-up alone.
The new £120 million for the UK AI Security Institute covers four financial years. Put both countries on a per-year basis and the gap is still large, just smaller than the headline.
Testing budget per year
Million US dollars · UK at about $1.33 per pound, the rate Semafor used
Why “10 times” needs a footnote
- Semafor: the £120m (~$160m) top-up “alone is more than 10 times” the US center’s annual budget
- But: the £120m is spread over four financial years, the US figure is one year
- Per year: the top-up is about $40m, roughly 3 to 4 times the US budget
- With AISI’s existing money: about 8 to 9 times
- Not counted: free compute and model access from companies, and testing by other agencies such as the NSA or the UK’s NCSC
Institute for Progress estimates: a “limited” US center needs $26m a year, an “equipped” one $84m. Matching the UK’s share of total government spending would mean about $850m a year.
Sources: HM Treasury, Budget 2026, 28 Oct 2026; AISI About page, accessed 29 Oct 2026; Institute for Progress, 13 May 2026; Semafor, 27 Oct 2026; House CJS report FY2027 (up to $15m). Per-year averages, ratios and dollar conversions are my arithmetic.
Powered by Thorsten Meyer AI
What More Testing Capacity Could Change
Government testing teams examine how advanced systems behave under risky conditions, including cybersecurity and biosecurity tasks. Their work can give policymakers and developers information about potential hazards before or after a model is released. The UK funding increase could support more evaluation work, technical standards and specialist capacity, though the Budget announcement does not set out a detailed spending plan or guarantee particular testing outcomes.
The figures also matter because they are being used to compare national approaches to AI oversight. A headline ratio based on the full UK four-year allocation and one year of US funding exaggerates the annual difference. Using annual figures makes the comparison clearer, while the US administration’s higher funding request shows that the gap could narrow if Congress approves it. Neither budget alone establishes how effective either country’s testing is.
For companies, public evaluation reports can offer practical information about observed model risks and testing methods. The source advises firms to check which rules apply to them and use published findings when assessing their own AI agents. For users, government testing should not be mistaken for certification or a guarantee of safety: it indicates that evaluators examined particular risks, not that every use or failure mode has been ruled out.
Products Worth Considering
Anker SOLIX S2000 Portable Power Station, 2,010Wh, 1,500W
Logitech MX Keys S Wireless Keyboard, Graphite
Bose QuietComfort Wireless Bluetooth Headphones with Active Noise Cancellation, Mic, USB‑C Charging, Deep Bass, 24‑Hour Battery – White Smok
As an affiliate, we earn on qualifying purchases.
How the UK and US Figures Compare
The Treasury first previewed the funding on 24 October, describing it as an investment in testing new AI systems and developing technical standards. The Budget document on 28 October confirmed the amount and specified the four financial years it covers. The source says the 2025 UK Spending Review had already set aside £240 million for AISI; that earlier figure is distinct from the newly announced £120 million.
The US counterpart operates within NIST. It was renamed the Center for AI Standards and Innovation in 2025, and the NIST page cited in the source now calls it the Center for Advancing Innovation and Standards for Super Intelligence (CAISSI). The centre describes itself as a government contact for industry testing, using voluntary agreements with developers and focusing on demonstrable risks such as cybersecurity, biosecurity and chemical weapons.
For fiscal 2026, the Institute for Progress, a Washington think tank, estimates the US centre had about $15 million: $10 million from appropriations and a $10 million Technology Modernization Fund loan spread across two years. The White House requested $27 million for fiscal 2027, while a House committee report provides for up to $15 million. A bipartisan proposal from Representatives Lori Trahan and Jay Obernolte would provide $100 million annually through 2029, but the source says none of those higher amounts has become law.
“£66m in funding per financial year”
— AI Security Institute
Funding Plans and Testing Access
The UK Budget gives the total and the four-year period, but not an annual breakdown or a detailed account of how AISI will spend the money. The US figures also include different kinds of funding and proposals: the $15 million estimate for fiscal 2026, the administration’s $27 million request for fiscal 2027, and legislative proposals are not interchangeable. Their final status may change as the US budget process proceeds.
The source reports that testing arrangements in both countries are voluntary. It also says Anthropic released Mythos 5.1 in September without giving access to organisations outside the US. The circumstances and implications of that access decision are not independently detailed in the supplied material. The source describes AISI test results involving other named models, but the figures are specific to those tests and do not establish how the systems would behave in every real-world setting.
Other uncertainties include the US centre’s staffing and leadership: the source says NIST does not publish a headcount and reports that director Chris Fall resigned in July, leaving no permanent director. These details may affect comparisons of institutional capacity, but the supplied material does not provide a comparable staffing breakdown for both countries.
How the New Budget Will Be Used
The next UK milestones are the allocation and deployment of the new funding across 2026-27 to 2029-30. The Budget announcement establishes the commitment, but the supplied material does not identify a timetable for new staff, specific evaluations or standards work. Further spending details will be needed to show how the annual capacity changes in practice.
In the US, Congress will determine whether the fiscal 2027 request or other proposals receive funding. The House committee’s provision and the bipartisan $100 million proposal remain distinct from enacted appropriations, according to the source. Until those decisions are made, the most defensible comparison is with the roughly $15 million fiscal 2026 estimate, clearly labelled as such.
For developers and organisations using AI agents, the near-term step is to follow public evaluation reports and check which regulatory requirements apply to their products and operations. More government testing may reveal additional risks, but voluntary access and limited public detail mean that the reach of the evaluations will remain an open question.
Key Questions
How much new money did the UK commit to AI testing?
The Budget confirmed £120 million for the AI Security Institute over four financial years, from 2026-27 to 2029-30. The source calculates an average of about £30 million a year; the Budget does not give a year-by-year split.
Is Britain spending more than ten times as much as the US?
Not on the annualised figures used in the source. It estimates UK funding, including existing AISI funding and the annual average of the new allocation, at about $128 million a year, against roughly $15 million for the US centre—around eight or nine times as much. Compared with the US administration’s $27 million fiscal 2027 request, the UK estimate is under five times as large.
What does the AI Security Institute test?
AISI evaluates advanced AI systems for risks, including potentially harmful capabilities. The source describes tests in which researchers examine how models behave under challenging conditions, but test results concern specific evaluations and do not amount to a general safety guarantee.
Is the US centre’s proposed higher budget approved?
No. The $27 million figure is a White House request for fiscal 2027, and the source says a House committee report provides up to $15 million. It also describes a separate bipartisan proposal for $100 million annually through 2029. The supplied material says none of those higher figures is law yet.
Does government testing mean an AI model is safe to use?
No. Testing means evaluators examined particular behaviours and risks; it does not certify that a model is safe in every setting or prevent every harmful outcome. The source says developers’ participation in testing is voluntary in both the UK and the US.
Source: ThorstenMeyerAI.com
