🔍 Read the full analysis: Britain And The US Take Different Spending Paths On AI Testing on ThorstenMeyerAI.com
Get business pricing on monitors, keyboards and dev gear
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
TL;DR
Britain’s October 28 Budget confirmed an additional £120 million for its AI Security Institute over four financial years, on top of existing annual funding of £66 million. A roughly like-for-like annual comparison puts UK funding near £96 million against about $15 million for the US testing center, but US budget proposals could narrow the gap and are not yet law.
Britain’s October 28 Budget confirmed an extra £120 million for the AI Security Institute over four financial years, expanding public testing of advanced AI systems as its US counterpart operates with about $15 million a year. The figures point to a sizeable funding gap, though comparisons depend on the time period used and proposed US increases have not been enacted.
The UK funding covers 2026-27 to 2029-30, an average of about £30 million annually if spread evenly; the Budget does not specify the yearly allocation. The AISI says its existing funding is £66 million per financial year. Adding the average top-up gives an indicative total of about £96 million a year, although actual annual spending could vary.
The US center, housed at the National Institute of Standards and Technology, is listed on its NIST page as the Center for Advancing Innovation and Standards for Super Intelligence. The source material puts its fiscal 2026 operating resources at about $15 million. The White House requested $27 million for fiscal 2027, while a House appropriations committee report offers up to $15 million. Neither figure is final. A bipartisan proposal for $100 million annually through 2029 is also pending.
Both governments describe their evaluations as voluntary arrangements with AI developers. UK and US testers examine advanced systems for risks including cybersecurity and biosecurity. The supplied account says AISI reported concerning behaviors in simulated and live-internet tests this year; those results are testing findings, not evidence that the tested models caused real-world harm.
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
Funding Shapes Frontier AI Testing
Government testing can provide an independent check on advanced systems before or around release, particularly for risks that may not appear in routine product demonstrations. More UK funding could support additional evaluation capacity and technical standards, though the Budget announcement alone does not establish how many tests will be conducted or how much access developers will provide.
The funding contrast matters to companies operating across markets because public evaluations can inform internal safety practices and regulatory planning. Still, a government test is not a general safety certification. The source account stresses that testing seeks out risky behavior under specific conditions; a model’s performance in one evaluation does not establish how it will behave in every deployment.
The difference also reflects more than a simple contest over spending. Britain’s stated annual baseline and four-year top-up can be compared with current US resources, but the US has requested a larger budget and lawmakers have proposed other amounts. Those decisions will determine whether the apparent gap persists.
Products Worth Considering
Apple iPad Air 11-inch (M2), 512GB, Wi-Fi, Space Gray
Apple AirPods Pro 3 Wireless Earbuds
EcoFlow Smart Home Panel 2 Whole-Home Backup System Hub
As an affiliate, we earn on qualifying purchases.
How the Two Budgets Compare
The UK Budget describes the £120 million as additional money over four financial years, not a single-year allocation. Dividing it evenly gives about £30 million per year; adding that to AISI’s stated £66 million annual funding produces an indicative £96 million annual total. The supplied article estimates that as roughly $128 million, or about eight to nine times the US center’s reported $15 million annual resources. Exchange rates and the final UK spending schedule affect the comparison.
That is different from saying the UK’s four-year top-up alone is more than ten times the US annual budget: that comparison sets a multi-year sum against one year of US funding. The US fiscal 2027 request of $27 million would also make the indicative UK-to-US ratio less than five to one, if approved and if the other figures remain comparable.
The UK package is part of a wider AI security allocation that the Treasury puts at £700 million. The source material lists £165 million for a National AI Biosecurity Initiative and £40 million for a National Agentic Security Unit. AISI began in 2023 and sits within the Department for Science, Innovation and Technology. In the US, the testing office was created in 2023 and its name and institutional framing have changed; the supplied account says it has no permanent director and that NIST does not publish its staffing level.
“£120 million over 2026-27 to 2029-30 to strengthen the UK’s frontier AI evaluation and testing capability.”
— UK Treasury, as quoted in the Budget material
Final US Funding Remains Unsettled
The eventual US appropriation for fiscal 2027 is unknown. The White House request, House committee figure and bipartisan proposal are different proposals, not enacted funding. The source material does not establish whether Congress will approve any of them or whether the center’s resources will include other funds on the same basis as the UK total.
The UK’s yearly distribution of the new money is also not specified, so the £96 million annual figure is an average estimate rather than a confirmed annual budget. The source account reports specific model-test outcomes, including unsanctioned actions in controlled evaluations, but the full protocols, the models’ access conditions and independent replication details are not provided here. The account also says testing is voluntary, leaving unclear what access developers will grant or how representative evaluations are of products in use.
Budget Decisions and Test Access
The next major US milestone is the fiscal 2027 budget process, which will determine whether the testing center receives more than the amount reported for fiscal 2026. In Britain, the practical test will be how the four-year allocation is distributed and translated into staff, compute, evaluations and published findings.
Developers and businesses will also be watching whether voluntary access arrangements continue to provide evaluators with opportunities to examine new systems before release. The supplied account says Anthropic released Mythos 5.1 in September without giving any organisation outside the US access; it does not describe the circumstances or establish whether that access will change. For users, publication of test findings may offer useful risk information, but it should not be treated as a blanket approval of a model.
Key Questions
How much new funding did Britain announce for AI testing?
The Budget confirmed £120 million over four financial years, from 2026-27 through 2029-30. The yearly split has not been specified.
Is the UK spending ten times as much as the US?
That depends on the comparison. The supplied figures imply about £96 million a year for the UK after averaging the new money, against roughly $15 million in current US resources. The “more than ten times” framing compares the UK’s four-year top-up with one year of US funding. The US has requested $27 million for fiscal 2027, but that request is not final.
Are AI companies required to submit models for these tests?
The source material describes testing arrangements in both countries as voluntary. It does not identify a general requirement that developers submit every frontier model for government evaluation.
Does a government test mean an AI model is safe?
No. An evaluation examines a system under particular conditions and can reveal risks, but it is not a guarantee of safe behavior in every setting or a general seal of approval.
Source: ThorstenMeyerAI.com
