🔍 Read the full analysis: Inside The Cost Of Checking AI Before It Ships on ThorstenMeyerAI.com
Get the latest gadgets delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
TL;DR
The UK 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 fair annualized comparison puts UK funding at about £96 million, against roughly $15 million for the comparable US testing center, though US budget proposals remain unsettled. Both countries rely on voluntary evaluations, and testing is not a government safety certification.
The UK Budget confirmed an extra £120 million for the AI Security Institute (AISI) to test frontier AI systems over four financial years, a funding boost aimed at expanding government evaluations of models before release. The institute already lists £66 million in annual funding; averaged across the new four-year allocation, the additional money amounts to about £30 million a year.
The Treasury’s October 24 preview described the investment as support for testing new AI systems and developing technical standards. The Budget document, published October 28, specified that the £120 million covers 2026-27 to 2029-30. It did not provide a year-by-year allocation, so £30 million annually is an average, not a disclosed spending schedule. Adding that average to AISI’s stated existing funding produces an approximate total of £96 million per year.
The funding is part of a wider £700 million AI security package, according to the Treasury. 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 will detect, investigate and respond to cases involving rogue AI agents. AISI, part of the Department for Science, Innovation and Technology, reports having more than 100 technical staff and priority access to over £1.5 billion in public computing resources.
The US comparison is less settled. The Center for Advancing Innovation and Standards for Super Intelligence (CAISSI), housed at the National Institute of Standards and Technology, has about $15 million for fiscal 2026, according to the Institute for Progress. The White House requested $27 million for fiscal 2027, while a House committee bill offers up to $15 million. Neither proposed figure is final law. A separate bipartisan proposal would provide $100 million annually through 2029, but it remains a proposal.
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 Pre-Release Testing
Government evaluations can give developers and the public evidence about how advanced systems behave under risky conditions, including cybersecurity and biosecurity scenarios. The UK’s additional funding could support more testing capacity, but the Budget announcement alone does not specify how many models will be assessed, which methods will be used, or what results will be made public.
The funding comparison matters because headline ratios can mislead. The £120 million UK allocation covers four years, while the US figure commonly cited is an annual amount. Comparing the UK’s existing £66 million with the £30 million annual average for its new funding gives roughly £96 million a year, or around $128 million at the source’s approximate conversion. Against the US center’s roughly $15 million fiscal 2026 funding, that is about eight to nine times; against the White House’s $27 million request, it is under five times. The exchange-rate conversion and the different status of funding figures make these approximate comparisons, not exact like-for-like measures.
For companies, these evaluations can inform internal risk reviews and preparedness for different regulatory systems. For users, a government test should not be read as an approval stamp: the source material describes evaluations designed to probe risks, not certify that a system is safe in every setting.
Popular right now
De'Longhi Magnifica Evo Fully Automatic Espresso Machine with Manual Milk Frother
Bose QuietComfort True Wireless Earbuds with Active Noise Cancellation, Black
Dyson Cyclone V10 Origin Cordless Vacuum
As an affiliate, we earn on qualifying purchases.
How UK and US Testing Works
AISI began in 2023 as the AI Safety Institute. The UK’s 2025 Spending Review had already allocated £240 million to the institute, and its website lists £66 million in funding per financial year. The new Budget money is therefore additional to that existing annual figure, rather than a replacement for it.
The US testing office also began in 2023 and has been renamed more than once. It was called the US AI Safety Institute and was renamed the Center for AI Standards and Innovation in 2025; the NIST page cited in the source material now uses the name CAISSI. NIST describes its role as a government contact point for industry testing, with voluntary agreements and evaluations focused on demonstrable risks such as cyberattacks, biosecurity and chemical weapons.
The source account also describes AISI evaluations involving model behavior in simulated or live environments, including reports of unsanctioned actions by AI agents. Those findings are reported test results, not evidence that every model will behave the same way outside those particular evaluations. The account says evaluations in both the UK and US are voluntary, a limit on what the funding alone can require from developers.
“£120 million over 2026-27 to 2029-30 to strengthen the UK’s frontier AI evaluation and testing capability.”
— UK Budget document
Testing Limits and Funding Gaps
The UK Budget does not say how the £120 million will be divided by year, how many staff or evaluations it will fund, or what reporting schedule will follow. The annualized £30 million figure is arithmetic based on an even spread, not a government-published annual budget.
US fiscal 2027 funding is unresolved: the president’s $27 million request and the House committee’s provision of up to $15 million are not enacted final funding figures. The source material also says the US center has no permanent director and that NIST does not publish its headcount. More broadly, voluntary testing means neither country’s evaluation office can be assumed to have access to every model or to assess every system before release.
Reported test findings require similar care. The source describes particular simulations and a live test, but the information provided here does not include full methodology, independent replication or enough detail to generalize results across models and uses. A government evaluation can identify risks under tested conditions; it cannot establish that a system is safe in all circumstances.
Budgets and Evaluations Ahead
The next funding milestone in the UK is the allocation of the new money across the 2026-27 to 2029-30 period and further detail on the evaluations it supports. The Budget announcement does not set out a public timetable for those decisions or name specific models for assessment.
In the United States, Congress must settle fiscal 2027 funding. The final amount will determine whether the center receives the White House’s requested $27 million, the House committee’s proposed amount of up to $15 million, or another figure. The separate bipartisan $100 million-a-year proposal would require further legislative action.
For now, companies can follow public evaluation reports and check which regulatory requirements apply to their systems, while treating outside testing as one input to their own safety work. The source material says the EU AI Office has legal authority to evaluate large models and fine their makers up to 3% of global turnover, adding a different regulatory approach to the voluntary UK and US systems. How that authority is applied in practice, and whether it leads to broader access to models for testing, remains to be seen.
Key Questions
How much new funding did the UK announce for AI testing?
The Budget confirmed £120 million over four financial years, from 2026-27 to 2029-30, for the AI Security Institute. The Budget did not publish an annual breakdown.
How does the UK amount compare with US funding?
AISI lists £66 million in existing annual funding. Averaging the new allocation over four years adds about £30 million annually, for roughly £96 million a year. The comparable US center has about $15 million for fiscal 2026, according to the Institute for Progress. Currency conversion and differences between enacted funding and proposals make the comparison approximate.
Does government testing mean an AI model is approved as safe?
No. The evaluations described are attempts to examine model behavior and identify risks under particular test conditions. They are not a general safety certification or a guarantee about every use.
Are UK and US companies required to submit models for testing?
The source material says testing by the UK and US institutes is voluntary. It does not establish that every developer provides access to every model before release.
What remains unsettled in the US?
Congress has not finalized the center’s fiscal 2027 funding. The White House requested $27 million, a House committee bill offers up to $15 million, and a separate bipartisan proposal seeks $100 million a year through 2029; these are not all enacted funding.
Source: ThorstenMeyerAI.com
![Musk Swaps Grok Parameters For A 4-Model AI Router [2026] – Https://tech-insider.org/](https://elfysworld.com/wp-content/uploads/2026/10/musk-swaps-grok-parameters-for-a-4-model-ai-router-2026-https-tech-insider-org-featured-560x300.jpg)
