The Alan Turing Institute
AI-generated text. This page was generated using artificial intelligence.
The Alan Turing Institute is the United Kingdom's national institute for data science and artificial intelligence. It describes itself as an independent charity, incorporated in England and Wales under company number 09512457 and charity number 1162533. Its AI-policy work includes research through the Centre for Emerging Technology and Security (CETaS), a centre within the institute concerned with technology and national security.[1][2]
Type: Research institute and registered charity
Policy focus: AI security, assurance, national resilience and surveillance governance
Official website: turing.ac.uk
Sources reviewed: October 10, 2026
Research and policy role
CETaS brings together researchers, technologists and practitioners, including expertise in law and ethics. Its stated mission is to strengthen UK security through research on emerging technologies. Its programme includes responsible AI use in cyber defence and intelligence analysis, evaluation of system reliability, and the effects of increasingly capable AI on state capacity and institutions. It also examines disinformation, democratic resilience and geopolitical competition.[2]
Surveillance governance is another part of this work: CETaS researches public attitudes to digital surveillance, privacy-impact assessment and proportionality in intelligence powers. The centre describes itself as non-political and says its research undergoes academic peer review and editorial review. Its publication policy distinguishes authors' views from those of the institute. These disclosures matter when interpreting an individual report's recommendations.[2]
AI sovereignty and resilience proposals
On September 2, 2026, CETaS published Sam Stockwell's Securing the UK's AI Future: A Sovereignty Vision for 2030. The briefing proposes combining selected domestic capabilities with diverse, trusted international suppliers. Its central concern is retaining practical control and fallback options across the AI supply chain. In his foreword, institute chief executive George Williamson connects this approach to national resilience.[3]
The proposed model portfolio combines frontier systems, open-weight models and domestic models suited to critical uses. It includes a medium-sized sovereign base model and stronger control over data provenance, training and assured access. For datasets, the briefing advocates federated, privacy-preserving arrangements for public-sector information, with attention to copyright, consent and licensing, as well as protection against poisoning and backdoors. For infrastructure, it calls for diversified compute and hardware supply to reduce dependence on individual suppliers.[3]
The briefing's 2030 vision also proposes a firmer institutional footing for the AI Security Institute, mandatory testing before deployment of high-risk frontier models, and a wider assurance and certification ecosystem. These are the briefing's policy recommendations. Its analysis of strengths, weaknesses, opportunities and threats spans models, hardware, compute and the wider research ecosystem, linking technical capacity to governance choices. The proposals are discussed in United Kingdom policy on catastrophic AI risk.[3]
Frontier-risk assurance
On September 23, 2026, the institute announced Frontier AI Risks: A practical way forward. Its announcement identifies cybersecurity, democratic instability, misalignment, loss of meaningful human control, and chemical or biological threats as five areas warranting practical action. It argues for verification and assurance of complete deployed systems, including the people, processes and organizations involved, alongside model evaluation.[4]
The institute's account emphasizes monitoring, safe override and fallback arrangements. It describes work alongside the AI Security Institute and National Cyber Security Centre. This contribution connects AI assurance to operational governance and institutional preparedness; the announcement presents a research agenda rather than reporting an enacted regulatory instrument.[4]
Disclosed programme funding
In a separate September 23, 2026 announcement, the institute reported that CETaS had secured a £2 million grant from Coefficient Giving for research on transformative AI's security risks. It described three priorities: changes to the security landscape, practical proposals for government preparedness, and resilience of societies and critical systems. The announcement identifies this support with the CETaS programme; it provides no institute-wide funding total.[5]
Related articles
- United Kingdom policy on catastrophic AI risk
- United Kingdom AI Law
- Organizations in AI Law and Policy
References
- ↑ The Alan Turing Institute, official website, institutional description and legal-entity footer, accessed October 10, 2026.
- ↑ 2.0 2.1 2.2 CETaS, About, accessed October 10, 2026.
- ↑ 3.0 3.1 3.2 Sam Stockwell, CETaS, Securing the UK's AI Future: A Sovereignty Vision for 2030, September 2, 2026, including George Williamson's foreword.
- ↑ 4.0 4.1 The Alan Turing Institute, Alan Turing Institute sets out practical agenda to tackle frontier AI risks, September 23, 2026.
- ↑ The Alan Turing Institute, Isabel Ryan, New £2 million grant to address security risks of transformative AI, September 23, 2026.