AI Safety Institute
This document outlines the establishment, mission, and functions of the UK AI Safety Institute (AISI), an evolution of the Frontier AI Taskforce. The Institute aims to minimise surprise from rapid advances in AI by conducting independent system evaluations, driving foundational safety research, and facilitating information exchange. Backed by long-term government investment, it will collaborate with international partners, industry, academia, and civil society to build an evidence base for safe AI development.
Summary and content fields are drawn from the document text and are pending editorial review.
Who issued it
Who it governs
This document is for policymakers, researchers, AI developers, and international partners interested in the UK's strategic approach to advanced AI safety and governance.
What the document contains
The document details the founding, mission, scope, core functions, and partnership frameworks of the UK AI Safety Institute, explaining how it will evaluate frontier AI models, perform safety research, and share information across government, industry, and academia.
The document does not provide binding regulatory rules, technical standards, legal penalties, or concrete pass/fail criteria for AI deployment.
- The UK Frontier AI Taskforce is transitioning into a permanent organisation known as the AI Safety Institute (AISI).
- The Institute's core functions are to conduct advanced AI system evaluations, drive foundational safety research, and facilitate information exchange.
- The AISI will focus on dual-use capabilities, societal impacts, system safety and security, and loss-of-control risks associated with frontier AI.
- The Institute is not a regulator itself, but its findings will inform domestic and international policymaking and regulatory frameworks.
- The Institute will operate with significant multi-year government funding and priority access to state-of-the-art compute resources like the AI Research Resource (AIRR).
· Ministerial foreword · Introduction · Mission and scope · Functions · Partnerships · Establishment · Definitions
How it scores
Each score is backed by a quote verified against the document text. Editorial review has not yet confirmed them.
Accountability
3/5Clear responsibility for AI outcomes
This mission stems from our conviction that governments have a key role to play in providing publicly accountable evaluations of AI systems and supporting research.
Mission and scope
screening score · 2026-10-23
Fairness
2/5Non-discrimination and equitable treatment
techniques to analyse training data for bias (see Box 2 ).
Mission and scope
screening score · 2026-10-23
Human Dignity
0/5Respect for the inherent worth of persons
screening score · 2026-10-23
Human Oversight
2/5Humans remain in control of consequential decisions
Abilities and tendencies that might lead to loss of control, such as deceiving human operators, autonomously replicating, and adapting to human attempts to intervene
Functions
screening score · 2026-10-23
Transparency
3/5Openness about how AI systems work and are used
To ensure that relevant parties receive the information they need to effectively respond to rapid progress in AI , the Institute will appropriately share its findings with policymakers
Facilitating information exchange
screening score · 2026-10-23
Enforcement
0/5Consequences and compliance mechanisms with real force
Functions
screening score · 2026-10-23