AI Wiki

ARIA (UK)

Research Organizations
27 min read
Updated
Suggest editHistoryTalk
RawGraph

Last edited

Fact-checked

In review queue

Sources

43 citations

Revision

v3 · 5,345 words

Fact-checks are independent of edits: a reviewer re-verifies the article against its sources and stamps the date. How we verify

The Advanced Research and Invention Agency (ARIA) is a United Kingdom government research funding body, established by Act of Parliament in 2022 and made operational in January 2023, that backs "high-risk, high-reward" science and technology which conventional grant systems are seen as too cautious to fund. Headquartered in London and sponsored by the Department for Science, Innovation and Technology (DSIT), it is modelled on the United States Defense Advanced Research Projects Agency (DARPA) and is widely called the "British DARPA." ARIA launched with a budget of approximately £800 million over four years, a figure the June 2025 Spending Review raised to at least £1 billion through the end of 2028-29. Its founding Chief Executive Officer was Ilan Gur, who led the agency from 2022 until the arrival of Kathleen Fisher, a former director of DARPA's Information Innovation Office, who became CEO in February 2026; the Chair is the technology entrepreneur Matt Clifford.[1][2][3][37][38]

ARIA must not be confused with two other bodies that share its name. The United States NIST ARIA programme (Assessing Risks and Impacts of AI) is an evaluation and benchmarking initiative run by the National Institute of Standards and Technology: the two share an acronym but differ fundamentally in purpose, since UK ARIA is a funding body that commissions speculative research across many scientific domains, while NIST ARIA is a sociotechnical testing programme aimed at measuring AI risk in deployment contexts. Separately, Meta's Project Aria is a line of egocentric research glasses built by Meta Reality Labs Research for multi-modal AI, robotics and machine-perception research; it shares the name but is a wearable hardware platform, not a funding agency. This article concerns the United Kingdom agency only.[4][5][43]

ARIA's portfolio spans neurotechnology, climate, robotics, plant biology, energy and computing, with several programmes that are directly relevant to artificial intelligence. The most prominent of these for the AI safety community is the Safeguarded AI programme, launched in April 2024 within the "Mathematics for Safe AI" opportunity space and backed by approximately £59 million in committed funding. It was designed and originally led by David "davidad" Dalrymple, with Nora Ammann taking over as Programme Director in April 2026. ARIA is institutionally distinct from the UK AI Security Institute (UK AISI), which sits inside DSIT and tests frontier AI systems for capability and risk, although both organisations are part of the same broader UK government science and technology architecture.[6][7][8][40]

How did ARIA come about?

The Cummings idea

The intellectual origin of ARIA lies in long-running advocacy by Dominic Cummings, a political adviser who became chief adviser to Prime Minister Boris Johnson in July 2019. Cummings had written in detail on his personal blog from 2017 onwards about the history of the United States Advanced Research Projects Agency (ARPA, later DARPA), and in particular about the period between 1962 and 1975 under directors such as J. C. R. Licklider, Robert Sproull and Stephen Lukasik, which he treated as a model of unusually productive state-funded science. Cummings argued that the United Kingdom's research funding ecosystem, then centred on UK Research and Innovation (UKRI) and its predecessor research councils, was risk-averse, peer-review-bound and structurally unable to support contrarian research bets.[9][10]

After Cummings entered Downing Street, the "ARPA project" became one of his publicly stated priorities. Press accounts at the time reported that his messaging-app profile summarised his agenda as "Get Brexit done, then ARPA." The Conservative manifesto of 2019 contained a commitment to "a new agency for high-risk, high-payoff research, at arm's length from government." Cummings left Downing Street in November 2020, but the idea survived his departure and was carried forward by the Department for Business, Energy and Industrial Strategy (BEIS) under successive ministers.[9][10]

From announcement to legislation

The agency was publicly named on 19 February 2021 by BEIS, then under Secretary of State Kwasi Kwarteng, with an initial budget allocation of £800 million over the spending review period. The Advanced Research and Invention Agency Bill was introduced to Parliament on 2 March 2021. After passing through both Houses and surviving a series of "ping-pong" amendments concerning freedom of information and accountability, the Bill received Royal Assent on 24 February 2022, becoming the Advanced Research and Invention Agency Act 2022.[11][12]

The Act creates ARIA as a statutory corporation with a small board, gives it a ten-year statutory protection from dissolution, exempts it from the Freedom of Information Act 2000, exempts its procurement from the standard public procurement rules, and assigns its sponsorship to the Secretary of State for Business, Energy and Industrial Strategy (later transferred to the Secretary of State for Science, Innovation and Technology when DSIT was created in February 2023). It also requires ARIA to have particular regard to scientific research and inventions that may bring about social, economic or environmental benefit.[11][13]

Becoming operational

Although the legislation passed in February 2022, the agency had no permanent leadership at that point. On 20 July 2022, BEIS announced the appointments of Ilan Gur as the first CEO and Matt Clifford as the first Chair. ARIA was formally established as an operational body on 26 January 2023, when its statutory functions came into force and the agency began hiring. The first cohort of Programme Directors, numbering eight, was announced in October 2023 and a second cohort of eight followed in April 2025. ARIA's first programme funding calls opened in early 2024.[1][2][14]

How is ARIA governed?

ARIA is a non-departmental public body, sponsored by DSIT through a framework agreement that sets out its accounting and reporting obligations. Its accountable officer is the CEO. The board is appointed by the Secretary of State and includes the CEO, the Chair and a small number of non-executive directors. Unusually for a British public body, ARIA is statutorily exempt from the Freedom of Information Act 2000 and from the Public Contracts Regulations, design choices justified by ministers on the basis that high-risk research requires confidentiality and procurement flexibility, and criticised by transparency campaigners and some parliamentarians during the Bill's passage.[11][12][15]

The Act also imposes a "tolerance to failure": ARIA is explicitly directed to expect that many of its bets will not work out, and the Secretary of State is required to have regard to that fact when reviewing its performance. ARIA is permitted to commission research, give grants, take equity, enter contracts and form subsidiaries, but it cannot itself confer degrees.[13]

How does ARIA's Programme Director model work?

ARIA's operating model is borrowed directly from DARPA and other ARPA-style agencies such as the United States ARPA-E (Advanced Research Projects Agency-Energy) and the more recent ARPA-Health (ARPA-H). The unit of execution is the Programme Director, an individual technologist or scientist hired on a fixed-term contract, typically three to five years, who is given an opportunity space and a budget and asked to design and run a single multi-million-pound programme of work. Programme Directors decide which teams to fund, how to structure milestones, and when to terminate projects.[14][16]

ARIA staff have publicly described their use of the Heilmeier Catechism, the set of nine questions popularised by former DARPA Director George Heilmeier in the mid-1970s, as a discipline that every Programme Director is expected to be able to answer for their programme. These questions include "What are you trying to do?" (in plain language), "How is it done today and what are the limits of current practice?", "What is new in your approach and why do you think it will be successful?", and "How will progress be measured?" The catechism is part of the conceptual core of the ARPA model that ARIA has imported.[17]

ARIA's funding instruments avoid traditional academic peer review. Programmes are designed by Programme Directors and approved internally; teams are selected by the Programme Director, often in collaboration with a small panel and with input from "Activation Partners," a network of nonprofits, research labs and venture firms that ARIA uses to source and support funded teams. This absence of peer review is one of the most significant departures from UKRI practice.[16][18]

Programme structure is typically broken into "Technical Areas" (TAs), themselves divided into work packages and phases with milestone gates. Funded teams are referred to as "Creators." A programme that begins with broad scoping work in a Phase 1 may be narrowed, expanded or terminated at the discretion of the Programme Director.[7][16]

Who leads ARIA?

Ilan Gur (founding CEO, 2022 to 2026)

Ilan Gur was announced as ARIA's founding CEO on 20 July 2022 and joined later that year. Gur holds a PhD in materials science from the University of California, Berkeley. He served as a Programme Director and Senior Adviser at ARPA-E between 2011 and 2014, then founded Cyclotron Road, a fellowship for science-based entrepreneurs hosted by Lawrence Berkeley National Laboratory, in 2014. Cyclotron Road's model was later spun out into the independent nonprofit Activate, of which Gur was founder and CEO until his appointment at ARIA. Gur's background combined direct experience of the ARPA model in the United States with experience of fellowship-style talent funding, both of which shaped ARIA's early operating choices.[19][20]

Gur announced in June 2025 that he would step down and return to the United States, describing the CEO post as one "designed to be time-bound" so that the agency stays "constantly renewed with fresh ambition."[42] He remained in post through the search and transition. ARIA received 314 applications for the role from "top researchers, successful founders, and ARPA veterans worldwide," named Kathleen Fisher as his successor in November 2025, and Gur returned to the United States once she took up the post in February 2026.[37][38]

Kathleen Fisher (CEO from 2026)

Kathleen Fisher took up the role of CEO in February 2026. A computer scientist, she created and led DARPA's High-Assurance Cyber Military Systems (HACMS) programme, later described by DARPA as one of its most influential of the decade and demonstrated by defending an autonomous helicopter against cyberattack, and she went on to lead DARPA's Information Innovation Office, overseeing a portfolio of more than 500 million dollars, over 50 programmes and more than 20 programme managers. ARIA called her "the standout candidate in an exceptional field." Her arrival further deepened ARIA's operational connection to the DARPA model.[37]

Matt Clifford (Chair)

Matt Clifford CBE has chaired ARIA since 2022. He is the co-founder of the talent investor Entrepreneur First. From late 2024 onwards he was also the United Kingdom government's independent reviewer for the AI Opportunities Action Plan, a strategy document commissioned by the Sunak government and published by the Starmer government in January 2025. Following publication, Clifford was appointed the Prime Minister's Adviser on Artificial Intelligence in 10 Downing Street, a role he held until stepping down in June 2025 for family reasons. He retained the chair of ARIA.[21][22]

Clifford's dual role placed ARIA at the centre of British AI policymaking during 2024 and 2025, since the Chair of the agency was simultaneously the principal external author of the government's AI strategy.[22]

How much funding does ARIA have?

ARIA's initial allocation was £800 million over the four-year period to 2025-26, announced as part of the 2020 Spending Review and confirmed in the agency's framework documents. Spending was deliberately back-loaded: ARIA had to be set up, staff hired, Programme Directors recruited and programmes designed before significant sums could be committed. By August 2025, reporting by Research Professional News found that ARIA had spent only about £16.5 million on research in its first two years, and was on course to leave close to £600 million of the original £800 million unspent by the end of the 2025-26 financial year.[23][38]

ARIA's leadership characterised this pace as intentional rather than a shortfall. CEO Ilan Gur told The Register that "we established the agency in January 2023 ... this is exactly as expected in terms of the timeline and scale," and Chair Matt Clifford, asked whether the Treasury had pressed the agency for savings, replied: "No ... We've not had any problems in this regard at all." The pattern reflects ARIA's design, in which large multi-year commitments are drawn down slowly: by late 2025 the agency reported having committed more than £400 million to the United Kingdom research and development ecosystem through the opportunity spaces launched in its first year, spread across programmes that run for three to five years.[37][38]

In the June 2025 Spending Review, the Treasury raised ARIA's budget to at least £1 billion and extended it to the end of 2028-29, reaffirming political support for the agency through to its statutory ten-year review; ARIA was also exempted from the government's request that departments identify at least 5 percent of savings ahead of the review. ARIA's annual accounts for 2024-25 reported around 53 staff and a year of operational spending of approximately £27.6 million.[14][23][38]

Individual programmes are sized in the range of £30 million to £100 million each. Examples include Precision Neurotechnologies (approximately £69 million), Massively Scalable Neurotechnologies (approximately £50 million), Safeguarded AI (approximately £59 million), Scaling Compute (approximately £50 million), and the climate cooling programme (approximately £57 million across a £50 million experimentation envelope and a separate £11 million scoping element).[7][14][24]

What is Safeguarded AI?

The most significant AI-focused programme in ARIA's portfolio is Safeguarded AI, which sits inside an opportunity space ARIA calls "Mathematics for Safe AI." The programme was designed and originally led by David "davidad" Dalrymple, who joined ARIA as a Programme Director in September 2023, and who in October 2023 published a programme thesis outlining the approach. ARIA officially opened the programme for proposals in April 2024 with a committed budget of approximately £59 million over a roughly three-to-five-year period.[6][25]

Programme thesis

The thesis of Safeguarded AI is that conventional approaches to AI safety, including reinforcement learning from human feedback, red-teaming and behavioural evaluation, do not provide the quantitative safety guarantees that society demands from safety-critical engineered systems such as civilian aviation or nuclear power. Safeguarded AI proposes to build a "gatekeeper": a verifier system that uses formal mathematical proofs against an explicit world model to certify, before deployment, that an autonomous AI system will not violate specified safety properties.[6][26]

Dalrymple drew the term "guaranteed safety" partly from earlier writings by AI researchers including Stuart Russell and Yoshua Bengio. In an influential 2024 position paper, Dalrymple, Bengio, Russell, Max Tegmark and other co-authors argued that future advanced AI systems should be constrained by mathematically verified contracts and that the path to credible safety lay in combining probabilistic world models with formal verification. Safeguarded AI is the practical research vehicle for that vision.[26]

Technical structure

ARIA originally divided Safeguarded AI into three Technical Areas:

  • TA1 (Scaffolding) funds work on building an extensible language and software platform for representing formal world models and for checking proof certificates that AI-produced artefacts satisfy specifications.[6][25]
  • TA2 (Machine Learning) funds work on using frontier machine learning to help domain experts construct best-in-class mathematical models of real-world dynamics and to train AI systems whose outputs can be formally verified.[6][25]
  • TA3 (Real-World Applications) funds deployment of verified autonomous AI in critical operating contexts, with a view to demonstrating quantitative safety guarantees in at least one domain.[6][25]

Yoshua Bengio, a Turing Award laureate and longstanding advocate of guaranteed-safety approaches, joined Safeguarded AI as Scientific Director in August 2024.[6][27]

Funded creator teams

By 2025, Safeguarded AI had funded a range of TA1 and TA2 teams across British and international academic institutions, startups and non-profits, including teams at the University of Oxford working on formally verifiable AI for business applications, and teams associated with research organisations such as Mila in Montreal, FAR AI and others. Programme materials emphasise that Safeguarded AI is interested in tool-building, not in deploying frontier general-purpose chatbots.[28]

Pivot and transition (2025 to 2026)

As frontier AI capabilities advanced faster than the programme had assumed, ARIA reshaped Safeguarded AI in a sequence of steps rather than a single pivot. In November 2025 ARIA expanded TA1, now framed as "Tooling," into a broader effort to build an open-source mathematical assurance and auditability toolkit, and it discontinued TA2 by cancelling its planned Phase 2 machine-learning solicitation. Dalrymple explained that rapid model progress had undercut the case for a dedicated machine-learning effort, writing that "it was more likely than not we'd select zero successful teams for Phase 2" and that "every frontier model has more capability than I expected at that point in time."[39]

In February 2026 ARIA refocused TA3, the applications area, onto cybersecurity, cancelling the planned Phase 2 call for cyber-physical applications while allowing existing Phase 1 projects to run to completion. ARIA said "frontier AI capabilities have advanced at a pace that has exceeded our original projections." A revised programme thesis published in 2026 sharpened the focus onto expanding TA1 "Tooling" and applying it to high-impact cybersecurity challenges, with cybersecurity and microelectronics named as the initial demonstration domains. In April 2026 Dalrymple transitioned from Programme Director to Technical Advisor, and Nora Ammann, who had helped run the programme as a Technical Specialist from before it began and who previously founded the research initiative Principles of Intelligence, became Programme Director. The approximately £59 million programme envelope was retained throughout.[39][40]

What other programmes does ARIA fund?

Safeguarded AI is one of around fifteen opportunity spaces that ARIA had publicly identified by early 2026. Opportunity spaces are framing categories: each may contain one or several discrete programmes run by individual Programme Directors. The following list, drawn from ARIA's website, summarises the main spaces and their associated programmes.[29]

  • Mathematics for Safe AI, hosting Safeguarded AI and a programme of smaller "opportunity seeds" in mathematical AI safety.[29]
  • Nature Computes Better, hosting the Scaling Compute programme, led by Suraj Bramhavar, which sets itself the goal of reducing the hardware cost of training large AI models by more than 1,000-fold. Bramhavar, a former co-founder and CTO of Sync Computing, argues that "modern AI algorithms offer an opportunity to exploit principles found in nature to dramatically reduce the cost of the underlying computational hardware," which he says could "fundamentally alter the existing supply chain" for AI. On 26 February 2026, ARIA joined CommonAI, a collaborative computing platform, to launch a Scaling Inference Lab embedded in operational data-centre environments, committing an initial £16 million as part of a £50 million investment.[30][31][41]
  • Scalable Neural Interfaces, hosting Precision Neurotechnologies (led by Jacques Carolan, around £69 million over four years, 18 Creator teams across non-invasive, remote, biological and adoption-focused workstreams) and Massively Scalable Neurotechnologies (around £50 million).[7][14][32]
  • Sculpting Innate Immunity, hosting the Sustained Viral Resilience programme.[29]
  • Bioenergetic Engineering, hosting Precision Mitochondria.[29]
  • Extending Our Perception, hosting Hypersensory Intelligence: Olfactory Perception, at the interface of AI sensor fusion and machine olfaction.[29]
  • Trust Everything, Everywhere, hosting the Scaling Trust programme on cryptographic and physical-digital trust infrastructure.[29]
  • Collective Flourishing, on modelling of human social dynamics.[29]
  • Scoping Our Planet, hosting Forecasting Tipping Points (led by Sarah Bohndiek and Gemma Bale, 27 international teams) and Enduring Atmospheric Platforms.[14][29]
  • Programmable Plants, hosting the Synthetic Plants programme.[29]
  • Engineering Ecosystem Resilience, hosting Accelerated Adaptation for biodiversity conservation.[29]
  • Future Proofing Our Climate and Weather, hosting Exploring Climate Cooling, the geoengineering programme led by Mark Symes.[24][29]
  • Smarter Robot Bodies, hosting Robot Dexterity, which intersects with AI through learned manipulation policies.[29][33]
  • Manufacturing Abundance, hosting Universal Fabricators, focused on atomically precise manufacturing.[29]

Of the listed programmes, Safeguarded AI, Scaling Compute, Hypersensory Intelligence and Robot Dexterity have the most direct AI content. Other programmes, such as Forecasting Tipping Points and Precision Neurotechnologies, rely on AI methods as components but are not classified as AI programmes.[7][29]

How does ARIA differ from DARPA?

ARIA is more closely modelled on DARPA than any other British research body. The similarities are deliberate: empowered Programme Directors with hire-and-fire authority over their portfolios, fixed-term tenures, an explicit "tolerance to failure," the Heilmeier Catechism as an internal discipline, the absence of academic peer review, the structuring of programmes into Technical Areas, and the rhetoric of "high-risk, high-reward" research. The appointment in 2026 of Kathleen Fisher, a former DARPA office director, as CEO tightened that lineage further.[16][17][37]

There are however four substantive differences. First, DARPA is part of the United States Department of Defense and exists primarily to deliver national-security capability; ARIA is not attached to any procurement customer. Critics, including the Wonkhe and War on the Rocks commentariat, have argued that this is the model's biggest weakness, since the disciplining effect of an end customer (such as the United States Air Force or Navy) is absent. Second, ARIA is much smaller than DARPA, both in absolute budget and as a share of national research spending. Third, ARIA includes climate, biology and neurotechnology programmes that DARPA does not run in comparable form. Fourth, ARIA's statutory FOI exemption is a stronger transparency carve-out than the equivalent exemptions DARPA enjoys under United States law, where DARPA contracts often eventually become subject to disclosure through Department of Defense reporting.[10][15]

ARIA staff have publicly argued that the appropriate comparators are not DARPA in isolation but the broader family of "ARPA-style" agencies, including ARPA-E, ARPA-H, the Defense Innovation Unit and Israel's chief scientist office.[16]

How is ARIA different from the UK AI Security Institute?

ARIA is part of, but distinct from, a constellation of United Kingdom government and government-adjacent bodies concerned with AI. The most important of the others is the UK AI Security Institute (UK AISI), known as AISI. AISI was created in November 2023 immediately after the AI Safety Summit at Bletchley Park as the AI Safety Institute, then renamed the AI Security Institute in February 2025. AISI sits inside DSIT, not at arm's length like ARIA. AISI conducts pre-deployment safety testing and red-teaming of frontier AI systems through agreements with leading AI developers including OpenAI, Anthropic and Google DeepMind, and runs a grants programme of more than £15 million to fund external AI safety research. AISI is part of an international network of AI safety institute bodies that includes the United States US AISI and equivalents in Singapore, Japan and elsewhere.[34][35]

The functional differences between ARIA and AISI can be summarised as follows. AISI is principally an evaluator of frontier AI systems and an advisor to government on AI risk; its in-house technical work tests existing systems against capability and misuse benchmarks. ARIA is principally a funder of speculative research and tools, with Safeguarded AI being the most prominent AI line of work. AISI's outputs are evaluations, reports and a small grant programme; ARIA's outputs are programmes, funded teams and (the agency hopes) deployable technologies. There is operational coordination between the two bodies, since both report to DSIT and Matt Clifford, ARIA's Chair, was the Prime Minister's Adviser on AI during 2025.[21][34]

Other adjacent institutions include UK Research and Innovation (UKRI), which remains the principal funder of mainstream British research; the Alan Turing Institute, the national institute for data science and AI, which is a partner organisation to several ARIA programmes; the Department for Science, Innovation and Technology itself, which sponsors ARIA and contains AISI; and several non-governmental AI safety bodies such as the Centre for the Governance of AI in Oxford. ARIA's place in this landscape is sui generis: it is the only United Kingdom body whose statutory remit is high-risk research with explicit tolerance to failure.[11][34]

What criticisms has ARIA faced?

ARIA was politically contested from the moment of its proposal. Concerns expressed during the Bill's passage included:

  • Lack of an end customer. Critics in Nature, Research Professional News and several House of Lords debates argued that DARPA's success relied on a clear procurement customer in the United States military and that ARIA, without one, would lack discipline. ARIA's response is that opportunity spaces themselves provide that discipline.[15][36]
  • Freedom of information exemption. Civil society groups and a cross-party group of parliamentarians objected to ARIA's exemption from the Freedom of Information Act 2000. The government argued that confidentiality was necessary for high-risk research; opponents argued that £800 million of public money should not be exempt from disclosure. The exemption survived intact.[11][12]
  • Relationship to UKRI. Academic commentators questioned how ARIA would relate to UKRI. The framework agreement clarifies that ARIA is additional to UKRI and not a substitute, but the boundary remains debated.[15]
  • Geoengineering. Exploring Climate Cooling, capitalised at approximately £57 million, has been the most publicly controversial ARIA programme. Critics including ETC Group, Geoengineering Monitor and individual climate scientists have argued that small-scale solar radiation modification experiments risk normalising large-scale geoengineering. ARIA has argued that the work is exploratory and that deployment decisions are for governments.[24][29]
  • Allocation of funds outside the United Kingdom. Reporting in 2025 found that a substantial fraction of ARIA's early commitments had flowed to United States-based teams. ARIA points to the international nature of frontier research and to the requirement that funded work create United Kingdom benefit.[23]
  • Slow spending. The finding that ARIA had spent only about £16.5 million of its £800 million by August 2025 drew scrutiny over whether the agency could deploy public money at scale. ARIA and its leadership argued that back-loaded, multi-year commitments were built into the model from the start.[23][38]
  • Performance measurement. Times Higher Education and others have asked how the public will know whether ARIA is succeeding given its long-horizon, high-failure-rate model. ARIA has published an evaluation framework emphasising programme-level rather than project-level success.[36]

Recent programmes and developments (2024 to 2026)

A non-exhaustive timeline of major announcements:

  • January 2024: ARIA published the Safeguarded AI programme thesis as a public document, setting out the case for guaranteed-safety AI research.[25]
  • April 2024: ARIA opened the Safeguarded AI call for proposals, with £59 million committed across three Technical Areas.[6]
  • April 2024: The Exploring Climate Cooling programme was announced.[24]
  • August 2024: Yoshua Bengio was announced as Scientific Director of Safeguarded AI.[6][27]
  • January 2025: Precision Neurotechnologies grants were awarded to 18 Creator teams, including teams at the University of Oxford and elsewhere; this was reported by Matt Clifford to have enabled the first National Health Service brain-computer interface clinical study.[14][32]
  • April 2025: The second cohort of eight ARIA Programme Directors took up post.[14]
  • April 2025: A £50 million envelope plus £11 million of preparatory funding for solar radiation modification experiments under Exploring Climate Cooling was confirmed.[14][24]
  • June 2025: Ilan Gur announced his departure as CEO. The Treasury raised ARIA's budget to at least £1 billion, extending its funding to the end of 2028-29.[14][23][38]
  • August 2025: Research Professional News reported that ARIA had spent about £16.5 million on research in its first two years, with the bulk of its budget still to be drawn down.[23][38]
  • November 2025: Kathleen Fisher, former Director of the DARPA Information Innovation Office, was announced as ARIA's next CEO. Safeguarded AI's first pivot expanded TA1 and cancelled TA2's Phase 2 machine-learning solicitation.[37][39]
  • February 2026: Fisher took up the CEO role. Safeguarded AI refocused TA3 onto cybersecurity, cancelling the planned cyber-physical Phase 2 call. ARIA joined CommonAI to launch a £50 million Scaling Inference Lab, with an initial £16 million grant, under the Scaling Compute programme.[31][40]
  • April 2026: David Dalrymple moved from Programme Director to Technical Advisor on Safeguarded AI and Nora Ammann became Programme Director; the revised programme thesis named cybersecurity and microelectronics as the initial demonstration domains.[40]

By early 2026, ARIA had committed more than £400 million across its opportunity spaces, with the bulk of programmatic spending still ahead under the renewed 2025 to 2029 envelope. Its long-term significance for British research and for global AI safety research will not be assessable until well into the 2030s.[37]

See also

References

  1. Advanced Research and Invention Agency. "About ARIA." https://www.aria.org.uk/ . Accessed 2026-05-19. ↩
  2. GOV.UK. "Advanced Research and Invention Agency." https://www.gov.uk/government/organisations/advanced-research-and-invention-agency . Accessed 2026-05-19. ↩
  3. Department for Business, Energy and Industrial Strategy. "Advanced Research and Invention Agency (ARIA): policy statement." https://www.gov.uk/government/publications/advanced-research-and-invention-agency-aria-statement-of-policy-intent/advanced-research-and-invention-agency-aria-policy-statement . Accessed 2026-05-19. ↩
  4. National Institute of Standards and Technology. "NIST Launches ARIA, a New Program to Advance Sociotechnical Testing and Evaluation for AI." 28 May 2024. https://www.nist.gov/news-events/news/2024/05/nist-launches-aria-new-program-advance-sociotechnical-testing-and . Accessed 2026-05-19. ↩
  5. NIST AI Challenges. "ARIA: Assessing Risks and Impacts of AI." https://ai-challenges.nist.gov/aria . Accessed 2026-05-19. ↩
  6. Advanced Research and Invention Agency. "Safeguarded AI programme." https://www.aria.org.uk/programme-safeguarded-ai/ . Accessed 2026-05-19. ↩
  7. Advanced Research and Invention Agency. "Precision Neurotechnologies." https://aria.org.uk/opportunity-spaces/scalable-neural-interfaces/precision-neurotechnologies/ . Accessed 2026-05-19. ↩
  8. AI Security Institute. "About." https://www.aisi.gov.uk/about . Accessed 2026-05-19. ↩
  9. Research Professional News. "Aria is an oldie, but there's no sign it will be a hit." 2021. https://www.researchprofessionalnews.com/rr-news-political-science-blog-2021-3-aria-is-an-oldie-but-there-s-no-sign-it-will-be-a-hit/ . Accessed 2026-05-19. ↩
  10. The Conversation. "ARPA: what is it and why does Dominic Cummings want one in the UK?" https://theconversation.com/arpa-what-is-it-and-why-does-dominic-cummings-want-one-in-the-uk-130975 . Accessed 2026-05-19. ↩
  11. UK Parliament. "Advanced Research and Invention Agency Act 2022." https://www.legislation.gov.uk/ukpga/2022/4/contents . Accessed 2026-05-19. ↩
  12. UK Parliament. "Advanced Research and Invention Agency Bill 2021-22: Lords amendments and 'ping-pong' stages." House of Commons Library briefing CBP-9433. https://commonslibrary.parliament.uk/research-briefings/cbp-9433/ . Accessed 2026-05-19. ↩
  13. UK Parliament. "Explanatory notes, Advanced Research and Invention Agency Act 2022." https://www.legislation.gov.uk/ukpga/2022/4/pdfs/ukpgaen_20220004_en.pdf . Accessed 2026-05-19. ↩
  14. Wikipedia. "Advanced Research and Invention Agency." https://en.wikipedia.org/wiki/Advanced_Research_and_Invention_Agency . Accessed 2026-05-19. ↩
  15. Nature. "UK 'DARPA' should let the sunshine in." 2022. https://www.nature.com/articles/d41586-022-00226-z . Accessed 2026-05-19. ↩
  16. Cambridge Industrial Innovation Policy. "ARIA and the value of challenge-led innovation." https://www.ciip.group.cam.ac.uk/reports-and-articles/aria-and-the-value-of-challenge-led-innovation/ . Accessed 2026-05-19. ↩
  17. DARPA. "The Heilmeier Catechism." https://www.darpa.mil/about/heilmeier-catechism . Accessed 2026-05-19. ↩
  18. Advanced Research and Invention Agency. "Funding opportunities." https://aria.org.uk/funding-opportunities/ . Accessed 2026-05-19. ↩
  19. Wikipedia. "Ilan Gur." https://en.wikipedia.org/wiki/Ilan_Gur . Accessed 2026-05-19. ↩
  20. Lawrence Berkeley National Laboratory. "Activate's Ilan Gur to Head UK Innovation Agency." https://elements.lbl.gov/news/activates-ilan-gur-to-head-uk-innovation-agency/ . Accessed 2026-05-19. ↩
  21. Wikipedia. "Matt Clifford." https://en.wikipedia.org/wiki/Matt_Clifford . Accessed 2026-05-19. ↩
  22. GOV.UK. "AI Opportunities Action Plan." January 2025. https://www.gov.uk/government/publications/ai-opportunities-action-plan . Accessed 2026-05-19. ↩
  23. Research Professional News. "Aria on track to spend a fraction of £800m budget by 2026." 2025. https://www.researchprofessionalnews.com/rr-news-uk-politics-2025-8-aria-on-track-to-spend-a-fraction-of-800m-budget-by-2026/ . Accessed 2026-05-19. ↩
  24. SRM360. "UK's ARIA Announces £45m Funding for 21 'Climate Cooling' Research Projects." https://srm360.org/news-reaction/aria-45m-funding-21-climate-cooling-research-projects/ . Accessed 2026-05-19. ↩
  25. Advanced Research and Invention Agency. "Safeguarded AI: constructing guaranteed safety." Programme Thesis V1, January 2024. https://www.aria.org.uk/wp-content/uploads/2024/01/ARIA-Safeguarded-AI-Programme-Thesis-V1.pdf . Accessed 2026-05-19. ↩
  26. The Next Web. "British DARPA to build AI gatekeepers for safety guarantees." https://thenextweb.com/news/british-darpa-aria-plans-ai-safety-gatekeepers . Accessed 2026-05-19. ↩
  27. Fortune. "Can AI be used to control safety critical systems? A UK-funded research program aims to find out." 10 June 2025. https://fortune.com/2025/06/10/aria-safeguarded-ai-safety-critical-systems-davidad-eye-on-ai/ . Accessed 2026-05-19. ↩
  28. University of Oxford, Department of Computer Science. "Oxford researchers awarded ARIA funding to develop safety-first AI for business." https://www.cs.ox.ac.uk/news/2437-full.html . Accessed 2026-05-19. ↩
  29. Advanced Research and Invention Agency. "Opportunity spaces." https://aria.org.uk/opportunity-spaces/ . Accessed 2026-05-19. ↩
  30. TechUK. "Aria announces submissions are open for new Scaling Compute - AI at 1/1000th the cost program." https://www.techuk.org/resource/aria-announces-submissions-are-open-for-new-scaling-compute-ai-at-1-1000th-the-cost-program.html . Accessed 2026-05-19. ↩
  31. PR Newswire. "ARIA joins CommonAI to launch Scaling Inference Lab, a £50 million investment to strengthen the UK's AI infrastructure." 26 February 2026. https://www.prnewswire.co.uk/news-releases/aria-joins-commonai-to-launch-scaling-inference-lab-a-50-million-investment-to-strengthen-the-uks-ai-infrastructure-302697637.html . Accessed 2026-05-19. ↩
  32. University of Oxford Institute of Biomedical Engineering. "IBME researchers part of teams receiving ARIA Precision Neurotechnologies grants." 28 January 2025. https://ibme.ox.ac.uk/2025/01/28/ibme-researchers-part-of-teams-receiving-aria-precision-neurotechnologies-grants/ . Accessed 2026-05-19. ↩
  33. Advanced Research and Invention Agency. "Robot Dexterity." https://aria.org.uk/opportunity-spaces/smarter-robot-bodies/robot-dexterity/ . Accessed 2026-05-19. ↩
  34. AI Security Institute. "Homepage." https://www.aisi.gov.uk/ . Accessed 2026-05-19. ↩
  35. TechUK. "Navigating the AI safety landscape: a comprehensive overview of the AI Safety Institute (AISI)." https://www.techuk.org/resource/navigating-the-ai-safety-landscape-a-comprehensive-overview-of-the-ai-safety-institute-aisi.html . Accessed 2026-05-19. ↩
  36. Times Higher Education. "Amid funding discord, how will we know if Aria is hitting the right notes?" https://www.timeshighereducation.com/depth/amid-funding-discord-how-will-we-know-if-aria-hitting-right-notes . Accessed 2026-05-19. ↩
  37. Advanced Research and Invention Agency. "Introducing ARIA's next CEO." November 2025. https://aria.org.uk/insights/introducing-our-next-ceo . Accessed 2026-07-14. ↩
  38. The Register. "UK science agency ARIA spared Treasury savings squeeze." 4 September 2025. https://www.theregister.com/2025/09/04/aria_spending_review/ . Accessed 2026-07-14. ↩
  39. ARIA Research (Substack). "November update: A Safeguarded AI pivot + real-time climate monitoring." November 2025. https://ariaresearch.substack.com/p/november-update-a-safeguarded-ai . Accessed 2026-07-14. ↩
  40. Advanced Research and Invention Agency. "Safeguarded AI." https://aria.org.uk/opportunity-spaces/mathematics-for-safe-ai/safeguarded-ai . Accessed 2026-07-14. ↩
  41. Advanced Research and Invention Agency. "Scaling Compute: AI at 1/1000th the cost." https://www.aria.org.uk/scaling-compute/ . Accessed 2026-07-14. ↩
  42. Research Professional News. "Ilan Gur to step down as Aria chief executive." June 2025. https://www.researchprofessionalnews.com/rr-news-uk-research-councils-2025-6-ilan-gur-steps-down-as-aria-chief-executive/ . Accessed 2026-07-14. ↩
  43. Engel, J., et al. (Meta Reality Labs Research). "Project Aria: A New Tool for Egocentric Multi-Modal AI Research." arXiv:2308.13561, August 2023. https://arxiv.org/abs/2308.13561 . Accessed 2026-07-14. ↩

Improve this article

Add missing citations, update stale details, or suggest a clearer explanation. Every suggestion is reviewed for sourcing before it goes live.

2 revisions by 1 contributor · full history

Suggest edit