British businesses supplying artificial intelligence products and services generated an estimated £33.6bn in AI-related revenue during 2025, an increase of 41%, as employment and investment across the industry expanded substantially.
The government’s Artificial Intelligence Sector Study 2025 estimates that 130,400 people were employed in AI-related roles, up 51% from the previous year. The industry’s estimated contribution to gross value added increased 58% to £18.6bn.
The research identified 10,488 active UK businesses providing AI products or services, compared with 5,862 in the preceding assessment. The increase reflects both expanding commercial activity and improvements in identifying companies operating in the sector.
Approximately 1,750 AI businesses were incorporated during 2025, while established companies also entered the research population as they developed relevant products or expanded their commercial activities.
Dedicated AI businesses accounted for 68% of companies identified, up from 56% in 2024. These organisations specialise primarily in AI-related products and services, rather than providing them alongside a wider range of unrelated commercial activities.
Their combined revenue almost doubled from £4.9bn to £9.2bn. However, large businesses continued to account for most of the industry’s economic activity, generating approximately 82% of total AI-related revenue.
Investment in dedicated AI companies reached £6.3bn during 2025, more than twice the previous year’s level. Much of that capital was concentrated in large transactions, with funding rounds worth at least £100m accounting for approximately 60% of investment value.
Two businesses attracted almost half the funding raised since the beginning of 2025, illustrating the concentration of capital among companies pursuing particularly resource-intensive technologies and infrastructure.
Smaller companies reported continuing difficulties accessing growth finance. Of the 376 AI businesses surveyed, 61% identified access to equity investment as a significant obstacle, while 57% said the availability of scaleup funding would materially affect their future activities.
Computing capacity is becoming another constraint. Some 72% of respondents reported an increase in computing requirements during the preceding year, while 91% of businesses using compute expected their costs to rise.
Energy prices and electricity grid connections were among the concerns affecting the provision of domestic AI infrastructure. These requirements are particularly substantial for businesses developing advanced models or operating large computing systems.
Regional concentration also remains pronounced. London, the South East, and the East of England accounted for 72% of registered office locations, although companies maintain additional operational sites elsewhere in the UK.
International investment has expanded alongside domestic activity. The research identified 1,372 internationally headquartered AI companies operating in Britain, equivalent to 13% of the sector, with more than half of those businesses headquartered in the United States.
The study also found that more than one in five UK-headquartered AI companies maintained at least one overseas location, indicating that international commercial relationships form an established part of the industry’s development.
The estimates combine financial accounts, commercial datasets, employment information, company surveys, and economic modelling. Revenue, employment, and gross value added are estimated where direct company figures are unavailable, while improvements in company identification affect comparisons with previous years.
Among businesses included in both the 2024 and 2025 assessments, AI-related employment increased by 31% on a like-for-like basis, providing a more consistent indication of expansion among established companies.
The findings describe a rapidly expanding UK AI industry, with revenue and employment growth accompanied by concentrated investment, demand for specialist skills, and increasing requirements for computing infrastructure.
You must be logged in to post a comment.