Establish an AI Safety Institute in the UK
“I can announce that we will establish the world’s first AI Safety Institute – right here in the UK.”
- Delivered
Where it stands
The AI Safety Institute was launched on 2 November 2023, at the AI Safety Summit at Bletchley Park, a week after the speech. It grew out of the Frontier AI Taskforce, with Ian Hogarth as its chair. On 14 February 2025 the next government renamed it the UK AI Security Institute and gave it a team on the criminal misuse of AI.
What has happened
- Oct 2023The Prime Minister says the UK will set up an AI Safety Institute, Promised
- Nov 2023The AI Safety Institute launches at the AI Safety Summit, with companies and countries including the US and Singapore agreeing to work with it, Delivered evidence
About the cost
No figure in the speech.
More detail
Sources (3)
How cards are checked
Status follows the promise standard, the same for every party. A status changes only on evidence. Anyone named on a card can reply, and replies are published next to it. AI Journalist is our automated reviewer: it re-reads every card against its live sources (the quote word for word, dates, evidence, status, cost, neutral wording, legal risk). An editor, Pavel Guzhikov, then checks each card against its sources and approves it.
Related promises
- Halve consultancy spending, saving £700m a yearLouise Haigh, LabourPromised
- Put AI centre stage in the G20 presidencyAndy Burnham, LabourPromised
- Give every mayor a share of income taxHM GovernmentPromised