UK minister contrasts £120 billion AI estimate with Google-funded £400 billion figure

Britain’s own central estimate of AI’s economic gains is less than a third of the £400 billion figure officials repeatedly promoted. In a parliamentary answer reported by WIRED on November 4, AI minister Kanishka Narayan put the government’s estimate at £120 billion by 2030 and traced the larger number to research commissioned by Google.

Narayan gave a range of £55 billion to £140 billion for the potential contribution of AI adoption and data investment. He called that assessment “hugely uncertain.” His answer did not withdraw the £400 billion figure or declare it wrong—a distinction that leaves the two estimates on the public record rather than replacing one with the other.

Adoption assumptions separate the forecasts

The larger estimate came from a publicly available 2023 report produced by consultancy Public First, now called PF Global. Narayan identified Google as its commissioner. Asked about the gap, PF Global’s head of economics, Jonathan Dupont, told WIRED that slower-than-expected adoption explained part of the difference.

Dupont said the government’s model assumed a much slower rate of take-up than the consultancy considered possible in 2023. PF Global said its own central estimate of the potential likely to be achieved by 2030 was now much closer to the government’s number, with growth accelerating in the years afterward.

The consultancy maintained that its view of AI’s overall potential had not changed. It pointed to more centrally directed adoption and integration into civil-service workflows as ways to improve take-up. Its explanation concerns how quickly potential gains can be realized, not simply whether the technology can produce economic benefits.

The higher number stayed in circulation

WIRED found that government departments and agencies had used the £400 billion figure for almost two years, across two governments. Ministers had the lower estimate by at least November 2025. Yet Skills England, the agency responsible for apprenticeships, was still publishing the older number four months later.

Conservative peer Lord Holmes of Richmond told WIRED that he had asked ministers about the £400 billion claim’s origins in January 2025 without receiving a response. He said the government should have clarified the figure sooner and more directly, stressing that decisions involving billions of pounds of public money depend on robust data.

Economic projects inevitably rely on assumptions, but overstating the benefits of technology risks undermining the case for change. The government needs to be realistic about what technology like AI can deliver and be transparent with the public about potential savings.

Dame Chi Onwurah, chair of the Commons Science, Innovation and Technology Committee, speaking to WIRED

Who checks the numbers behind policy?

Onwurah supported promoting AI’s potential to improve public services, but identified hype as a barrier to digital transformation. Her criticism was about how the case for change is presented: inflated expectations, she warned, could weaken support for the improvements government wants to make.

Sumedha Deshmukh of the Ada Lovelace Institute called for the government to test economic claims before using them to build policy. In comments to WIRED, she urged ministers to fund independent research and measure whether promised gains actually arrive. Apparently precise AI forecasts, she argued, can contain assumptions and incentives shaped by the industry funding the work.

PF Global defended the independence of its research, saying all judgments and assumptions were its own even as its work for Google continued. That defense addresses who controls the analysis. Deshmukh’s proposed test goes further: establish where a number comes from, what it assumes and who benefits before government relies on it.