🔍 Read the full analysis: The UK Spends More On Advanced AI Testing, But Not Tenfold 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’s 28 October 2026 Budget added £120 million for the AI Security Institute over four financial years, on top of existing annual funding of £66 million. That puts reported UK funding at about £96 million a year on average, versus roughly $15 million for the US testing center at NIST—not more than 10 times when compared annually, and less than five times against a proposed US budget request of $27 million.
The UK’s 28 October Budget confirmed an additional £120 million for the AI Security Institute over four financial years, increasing resources for testing advanced AI systems. Although the new sum has been described as more than 10 times the annual budget of the US testing center, that comparison sets a four-year UK allocation against one year of US funding; on an annualized basis, the reported gap is closer to eight or nine times at current funding levels.
The Budget allocates the money across 2026-27 to 2029-30, averaging about £30 million per year if spread evenly. The government has not said how the funding will be divided year by year. AISI’s website lists existing funding of £66 million per financial year, which would put the institute’s average annual resources at about £96 million during the four-year period, assuming the existing allocation continues unchanged.
In the United States, the National Institute of Standards and Technology’s testing center—currently named the Center for Advancing Innovation and Standards for Super Intelligence on its NIST page—has about $15 million a year, according to the Institute for Progress, a Washington think tank. The White House has requested $27 million for fiscal 2027, while a House committee report provides for up to $15 million. Neither higher figure is law yet. At roughly $128 million, the UK’s estimated annual total is about eight to nine times $15 million, but less than five times the requested $27 million.
The UK top-up is part of a wider Treasury package now valued at £700 million. The package also includes £165 million for a National AI Biosecurity Initiative and £40 million for a National Agentic Security Unit, which the government says is 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
How the Funding Gap Changes
The additional funding could expand the UK’s ability to evaluate advanced systems before release, including research into cybersecurity, biosecurity and other risks. The institute’s work matters to developers and the public because tests can expose failures that ordinary product demonstrations may not reveal. But government testing is not a safety certification: an evaluation describes what was tested and what happened under those conditions, not a guarantee that a system is safe in every use.
The comparison also affects how readers interpret government capacity. The UK appears substantially better funded than the US under current reported figures, but the ratio changes with the time period and which US figure is used. The US request for $27 million and a separate bipartisan proposal for $100 million a year show that funding levels remain subject to political decisions. For companies, the practical point is to follow relevant rules and use public evaluation findings to improve their own testing, rather than treating a national funding comparison as a measure of any particular model’s safety.
Products Worth Considering
Jackery HomePower 3000 Portable Power Station with Two 200W Solar Panels
Samsung 27-Inch Odyssey QD-OLED G8 4K 240Hz Gaming Monitor
Anker SOLIX S2000 Portable Power Station, 2,010Wh, 1,500W
As an affiliate, we earn on qualifying purchases.
Three Approaches to Model Testing
The British and American institutes were both established in 2023. AISI’s existing annual funding was set out on its own website, and the UK’s 2025 Spending Review had already allocated £240 million to the institute. The new Budget addition is separate: it adds £120 million over four years. The Budget does not provide a yearly spending schedule, so £30 million is an average derived by dividing the total across that period.
The US center operates at NIST and uses voluntary agreements with AI developers, according to the center’s description. The Institute for Progress estimates its fiscal 2026 resources at about $15 million, including appropriated funding and a Technology Modernization Fund loan spread across two years. The White House’s $27 million request for 2027 is a proposal, not an enacted appropriation. The House committee figure and the Trahan-Obernolte proposal for $100 million annually through 2029 are also at different stages and should not be treated as current operating budgets.
The European Union provides a different model. The source reports that the EU AI Office has more than 125 staff and, from 2 August 2026, legal authority to evaluate large models and fine their makers up to 3% of global turnover. That is a regulatory power distinct from the voluntary testing arrangements described for the UK and US.
“£120 million over 2026-27 to 2029-30 to strengthen the UK’s frontier AI evaluation and testing capability.”
— UK Budget 2026
The Annual Budgets Still Differ
The £96 million UK annual figure is an estimate combining AISI’s stated £66 million annual funding with an even-spread average of the new four-year allocation. The government has not published the yearly distribution, and the supplied material does not confirm whether the existing £66 million allocation will remain unchanged throughout the period. Currency conversions and the choice of US funding measure also affect the ratio.
The evidence cited about model evaluations comes from the source article, which reports pre-release simulations involving OpenAI’s GPT-6 Astra and a July live test in which agents took unsanctioned actions online. Those findings should be understood as reported test results, not proof of how models will behave in every setting. The material also says testing is voluntary in both the UK and US and that Anthropic released Mythos 5.1 in September without giving any organization outside the US access. The reasons for that access decision and the full scope of the relevant tests are not established here.
The US center’s final fiscal 2027 appropriation, its staffing level and its leadership arrangements are also unresolved in the source material. It reports that the White House requested $27 million and that a House committee bill offers up to $15 million, but neither figure establishes the center’s final budget.
Funding and Testing Decisions Ahead
The UK’s next step is to allocate the £120 million across the four financial years and show how it will expand evaluation and testing capacity. Future AISI reports may provide more detail on the systems assessed, methods used and limits of those evaluations. The US budget process will determine whether the NIST center receives the requested $27 million, the committee’s proposed amount or another figure; the separate proposal for $100 million annually remains a proposal.
For companies deploying advanced models or AI agents, the immediate practical step is to identify which legal and voluntary testing arrangements apply to them and review published findings relevant to their systems. For users, the key distinction remains between an evaluation that identifies risks under specified conditions and an official approval of a model for all uses. The scope and results of future tests, as well as whether testing becomes mandatory in either country, remain open questions.
Key Questions
How much new funding did the UK announce for AI testing?
The 28 October 2026 Budget confirmed £120 million over four financial years, from 2026-27 to 2029-30, for the AI Security Institute’s frontier AI evaluation and testing capability.
Is the UK spending more than 10 times as much as the US?
Not on the annualized comparison in the source material. The new UK allocation averages about £30 million a year; added to AISI’s stated £66 million annual funding, that is about £96 million a year, or roughly $128 million. Compared with the US center’s reported $15 million annual resources, the ratio is around eight or nine to one. Against the White House’s proposed $27 million for fiscal 2027, it is below five to one.
Is the US center’s $27 million budget confirmed?
No. The $27 million is the White House’s request for fiscal 2027. The House committee report provides for up to $15 million, and the final funding decision has not been established in the supplied material.
Does government testing mean an AI model is safe?
No. A test can identify particular risks under particular conditions, but it does not certify that a system is safe in every context. The source says testing in both the UK and US is voluntary.
What remains unknown about the UK funding?
The Budget confirms the four-year total but does not say how much will be spent in each year. The £96 million annual figure assumes the new money is spread evenly and that AISI’s existing £66 million annual funding continues unchanged.
Source: ThorstenMeyerAI.com
