UK sovereign AI fund launches with £500m for domestic computing

TL;DR: The UK’s sovereign AI fund, backed by £500 million from the Department for Science, Innovation and Technology, formally launches on 16 April. The fund will build domestic computing infrastructure to reduce dependence on foreign cloud providers and keep AI intellectual property within UK borders.

The UK government is putting £500 million behind a new sovereign AI fund designed to keep the country’s growing AI sector running on home-grown infrastructure. Chaired by James Wise, Partner at Balderton Capital, the fund formally launches on 16 April and will coordinate efforts across investors, industry leaders, and public agencies.

Reducing foreign dependency

The fund addresses a practical problem: UK businesses storing sensitive data on AWS, Google Cloud, or Microsoft Azure must navigate complex legal frameworks around cross-border data flows. By expanding domestic supercomputing access through facilities like Isambard-AI in Bristol and Dawn in Cambridge, the government aims to offer local alternatives with lower latency and simpler regulatory compliance.

An initial £8 million in seed capital has already gone to the OpenBind Consortium, which maps molecular binding at a scale 20 times larger than any previous database. For pharmaceutical companies, this domestic dataset could cut drug discovery timelines and reduce research costs by up to 40%.

Hardware and talent investment

The government has committed up to £100 million in Advance Market Commitments, acting as a first customer for domestic hardware developers. New Growth Zones in South Wales and Culham will provide the physical space and power needed for data centre expansion.

Talent remains a bottleneck. The fund is expanding the Encode fellowship to attract top global researchers into UK laboratories, building a pipeline of engineers who can work with domestic hardware architectures.

Looking forward

The UK’s tech sector already supports a £1 trillion market with more than 200 unicorns and over 5,800 AI companies, the largest in Europe. Whether this public investment can meaningfully shift the balance away from commercial hyperscalers will depend on adoption rates and whether domestic infrastructure can match the scale and reliability that enterprises expect.