Civil servants in at least eight government departments and the police are using artificial intelligence to make decisions on welfare, immigration and criminal justice, it has been claimed.
The Guardian reported that the systems were being used in the Department for Work and Pensions for decisions on benefits, that facial recognition software was used by the Metropolitan Police, and an algorithm was used by the Home Office to spot sham marriages.
Rishi Sunak has previously said he believes AI could improve public services if used safely and securely. Michelle Donelan, the science secretary, will give a speech today on AI strategy and how to ensure security.
Next week the government will hold an AI safety summit at Bletchley Park, Milton Keynes. No 10 has declined to publish a list of world leaders attending but the prime minister's spokesman said: "We remain confident that we will get the right people there, both in terms of world leaders but also leading tech firms, leading businesses and experts in the AI space."
He added "no one was looking at this" anywhere else in the world, and said: "So it's worth acknowledging that this will be the first time that there has been any gathering to consider the issue of AI safety, and that in itself is hugely worthwhile. We remain confident that we will have the right group of individuals to look carefully at this issue in terms of how we set the right guardrails so we can maximise the opportunities."
Sunak previously said: "Obviously there are benefits from artificial intelligence for growing our economy, for transforming our society, improving public services. But ... that has to be done safely and securely and with guard rails in place, and that has been our regulatory approach . . . so that we can exploit AI for its benefits."