Policy / Organization profile

UK Government

Organization overview

What UK Government does

The AI Security Institute (AISI) is a UK government research organisation within the Department for Science, Innovation and Technology (DSIT). Its mission is to equip governments with a scientific understanding of risks posed by advanced (frontier) AI, by conducting and commissioning technical research, running evaluations, and developing/validating mitigations that can inform government governance and international collaboration. AISI was established as the evolution of the UK’s Frontier AI Taskforce (announced in April 2023) and was launched at the 2023 AI Safety Summit; the institute’s policy/technical scope is designed to connect frontier AI evidence to both domestic and international AI governance work.

Organization monitoring

News

Company record

Sep 23, 2026 · Official · GOV.UKDr Jade Leung appointed Vice-Chair of the AI Security Institute and Security Advisor to the AI Taskforce

The UK government announced Jade Leung’s appointment as Vice-Chair of AISI and a security advisory role to the AI Taskforce, alongside a transition away from her Prime Minister’s Adviser AI role at the end of September 2026.

↗

2026-06-?? · Official · GOV.UKMemorandum of Understanding between UK and ElevenLabs on AI opportunities

A UK government publication describes a MoU between the UK and ElevenLabs, including continuation/deepening of the research partnership with the UK AI Security Institute (AISI) established in February 2026.

↗

Jan 29, 2026 · Official · GOV.UKSecure AI infrastructure: call for information

A government notice invites input to help understand capabilities and constraints in developing secure AI computing systems, coordinated by DSIT, AISI, and the National Cyber Security Centre.

↗

Dec 11, 2025 · Official · GOV.UKAI to accelerate national renewal and growth as Google DeepMind backs UK tech and science sectors

A GOV.UK press release describes a partnership with Google DeepMind and highlights expanded collaboration with the UK AI Security Institute to support safe and responsible AI development.

↗

Dec 11, 2025 · Official · GOV.UKMemorandum of Understanding between the UK and Google DeepMind on AI opportunities and security

A published MoU states that Google DeepMind will provide AISI with priority technical access to frontier models and outlines joint research areas related to AI safety, security, and societal resilience.

↗

Mar 05, 2025 · Official · GOV.UKCutting-Edge Research on AI Security bolstered with new Challenge Fund to ramp up public trust and adoption

The UK government announced AISI’s first grant fund (Challenge Fund) and provides programme details, including total programme funding and grant size.

↗

Date not disclosed · Official · GOV.UKInaugural report pioneered by AI Security Institute gives clearest picture yet of capabilities of most advanced AI

GOV.UK reports on an inaugural AISI report describing evidence from controlled tests of advanced AI capabilities and related implications for safety and security evaluation.

↗

Source map · 2 recurring channels · 21 references

Still resolving: Newsroom · Official X

Capital

Funding

Latest disclosed valuationNot disclosed

Tracked capital$345M4 sourced rounds

DateRoundRaisedValuationLead / investorsEvidence

UndatedGovernment funding allocation (Spending Review 2025)$240M—

Not attributed

Official ↗

UndatedGovernment grant programme (Challenge Fund)$5M—

Not attributed

Official ↗

UndatedGovernment investment for the predecessor (Frontier AI Taskforce)$100M—

Not attributed

Official ↗

Undatedclarification_needed——

Not attributed

Official ↗

Products & focus

What it builds

Inspect

Inspect is an open-source framework for frontier AI evaluations developed by the UK AI Security Institute (UK AISI) (with Meridian Labs mentioned in AISI materials). It provides composable evaluation components, large collections of pre-built evaluations, evaluation tooling (including a web-based Inspect View tool and a VS Code extension), and sandboxing mechanisms for running untrusted model code.

source ↗

Challenge Fund (grant programme)

AISI runs the Challenge Fund grant programme to support external research teams with awards of up to £200,000 per project on AI safety and security topics.

source ↗

People

Leadership

Ian HogarthChair Henry de ZoeteDirector Jade LeungChief Technology Officer Nate BurnikellChief Strategy Officer Chris SummerfieldResearch Director

Operating signals

Key metrics

Funding per financial year£66m (plus long-term resourcing commitments)Grant funding capacity>£15 million (in grant funding)Compute / access (priority)>£1.5 billion of compute in the UK’s AI Research Resource and exascale supercomputing programme (as stated)Staff scaleOver 100 researchers (as stated in the AI Opportunities Action Plan update)Jan 01, 2025 Frontier models tested30 frontier models tested (as stated in the AI Opportunities Action Plan update)Jan 01, 2025 Open-source evaluation outputAISI open-source Inspect platform is described as an international reference point for AI evaluationsJan 01, 2026

Company history

Milestones & partnerships

Jan 29, 2026

Secure AI infrastructure information-gathering initiative

DSIT, AISI, and the National Cyber Security Centre issued a call for information to gather views relevant to secure AI computing systems research and technical pilots.

↗Dec 11, 2025

MoU with Google DeepMind on AI opportunities and security

AISI and DSIT entered a published MoU framework with Google DeepMind, including priority technical access to frontier models for AISI under the MoU period.

Google DeepMind

↗Mar 05, 2025

Challenge Fund launched as the institute’s first grant fund

The government announced the first grant fund created under AISI (Challenge Fund) with grants up to £200,000 and programme focus areas in AI security research.

↗2024-05-??

Open-sourcing evaluation tooling (Inspect)

AISI published an open-sourcing post describing Inspect as an open-source framework for frontier AI evaluations.

↗Nov 01, 2023

Launch at the Bletchley Park AI Safety Summit (as AI Safety Institute)

AISI’s careers page states it was launched at the Bletchley Park AI Safety Summit in 2023.

↗Date not disclosed

Established as successor to the Frontier AI Taskforce

The AI Safety Institute was described as an evolution of the UK’s Frontier AI Taskforce, becoming a permanent feature of the AI ecosystem as of the policy publication timeline.

↗