The UK’s AI Windfall Comes With a Much Smaller Number

For two years, the UK government told the public that artificial intelligence could add £400 billion to the economy. That figure helped present AI as a source of growth at a time when many people feared sweeping job losses and even an economic crash.

Now the government has put a much smaller number beside it. AI minister Kanishka Narayan said AI adoption and data investment could add between £55 billion and £140 billion to the UK economy by 2030, with a central estimate of £120 billion. He described that newer assessment as a “hugely uncertain” estimate.

Where the £400 billion figure came from

Narayan gave the figures in response to a parliamentary question. He also explained that the £400 billion estimate “was drawn from a publicly available 2023 report commissioned by Google and produced by Public First,” a British consultancy firm now referred to as PF Global.

That distinction matters because government departments and agencies used the £400 billion figure for almost two years, under two governments. Former technology minister Peter Kyle also used it while promoting the benefits of AI, and Skills England, the government agency responsible for apprenticeships, was still publishing the old figure four months after ministers had the lower estimate.

Ministers have had the lower estimate since at least November 2025. Lord Holmes of Richmond had already asked ministers about the origins of the £400 billion claim in January 2025, but he did not receive a response.

Narayan’s answer is the first recorded evidence of the government setting the two numbers side by side and saying that the larger figure came from research commissioned by Google. The answer did not withdraw the £400 billion figure or say it had been wrong.

Different assumptions, different outcomes

PF Global’s head of economics, Jonathan Dupont, said the gap was partly linked to the government’s slower adoption of AI. “AI is a very fast moving technology,” Dupont said, adding that the government’s figures assume “a much slower rate of adoption than looked possible back in our model that was created in 2023.”

PF Global said its view of AI’s overall potential has not changed. The difference comes from what has happened in practice since the earlier model was created. With adoption moving more slowly, the consultancy said its central estimate for the amount of potential likely to be achieved by 2030 is now much closer to the government’s figure, with growth ramping up in the next few years after that.

PF Global also pointed to more top-down adoption of AI and the integration of AI into civil service workflows as ways for take-up to improve. In other words, the size of the eventual benefit depends not only on what the technology can do, but also on how quickly government departments put it to work.

That helps explain how two large numbers can exist without describing the same thing. The £400 billion figure reflected one set of assumptions about adoption and future use. The £120 billion central estimate reflects a newer assessment shaped by the slower rate of adoption seen so far.

The wider argument over AI claims

Dame Chi Onwurah, who chairs the Commons Science, Innovation and Technology Committee, said the government is right to promote AI’s potential to improve public services. But she warned that the figures also show one of the main barriers to delivering digital transformation: hype.

“Economic projects inevitably rely on assumptions, but overstating the benefits of technology risks undermining the case for change,” Onwurah said. She added that the government needs to be realistic about what AI can deliver and transparent with the public about potential savings.

Lord Holmes also criticised the delay in clarifying the number. “It is unfortunate that the figure used repeatedly by the government, and then, unsurprisingly, by multiple other organizations has now been significantly and rightly adjusted,” he said. “It would have been better for the government to clarify this much sooner and more directly.”

He added: “Robust data is critical, not least as billions of pounds of our money and how the government chooses to deploy it depend on it.”

Sumedha Deshmukh of the Ada Lovelace Institute welcomed the revised estimates but said it was concerning that they came after almost two years of ministers citing research paid for by Google. She described the episode as part of a wider pattern in which AI figures can look precise even though they contain assumptions and incentives, with industry funding or shaping some of the research behind them.

Deshmukh said the government should test the numbers it receives before building policy on them, fund independent research, and measure whether promised gains arrive. Before using any figure about AI’s economic or productivity benefits, she said officials should ask three basic questions: where did it come from, what does it assume, and who benefits?

The dispute between £120 billion and £400 billion is therefore about more than a single forecast. It raises a basic question about how governments should talk about new technology: how do they support useful change without turning uncertain possibilities into promises?

Stay connected via Google News

Follow us for the latest AI News