Britain’s AI sovereignty plan built on foreign foundations

CYBER

Heavy dependence on US technology, data centres and supply chains raises questions over whether the UK can ever control its digital future

By Denise Schipani, Cyber and Geopolitical Intelligence Analyst

BRITAIN’S new Sovereign AI Fund will fail to secure digital sovereignty, with continued reliance on US technology leaving national security exposed.

The warning follows the launch last week of the UK’s Sovereign AI Fund, in which Tech Secretary Liz Kendall pledged to reduce dependence on foreign innovation and restore control over Britain’s technological future.

A key component of the AI Opportunities Action Plan launched in January 2025, the fund will invest £500 million in up to 30 domestic AI startups, aiming to boost innovation and drive economic growth.

But US firms continue to dominate the sector. Companies including OpenAI and Anthropic lead the generative AI market in 2026, with combined venture funding reportedly reaching $242.6 billion – far outstripping UK competitors.

The UK remains a top-tier player, ranking behind only the US and China in AI investment and development. Its AI market is projected to exceed £1 trillion by 2035, with strong performance in patents and quantum research.

Yet much of the technology underpinning that growth is foreign-owned.

US-developed systems – including those from OpenAI, Anthropic and Google’s Gemini – are already deeply embedded in the UK economy, with 57% of organisations planning to adopt AI within the next three years.

The Sovereign AI Fund aims to counter this by giving selected UK firms access to national supercomputing capacity, alongside government support on procurement, product validation and visas for skilled workers.

Kendall urged the public to “make AI work for Britain”, calling domestic capability “critical for our economic prosperity and national security”. James Wise, chair of the Sovereign AI Unit, says the initiative could be “transformational”.

However, analysts warn that funding alone will not deliver sovereignty without control over underlying infrastructure.

Concerns centre on the UK’s limited data centre capacity – a critical constraint on AI development.

While Prime Minister Keir Starmer announced plans in January 2025 for a major AI data centre in Loughton, construction had yet to begin as of March 2026. With typical build times of 18 to 36 months, the project is unlikely to meet its end-2026 target.

Efforts to accelerate construction have also triggered resistance from local authorities over environmental concerns, including energy consumption, water use and biodiversity risks.

A recent report by the Open Rights Group underlines the scale of the challenge, and finds the UK remains heavily dependent on a small number of US hyperscalers – global tech firms that provide vast computing capacity through large-scale data centres to businesses looking to scale.

With the US hosting more than half of global hyperscale infrastructure, its ability to restrict access through sanctions poses a potential risk to any sovereignty strategy.

That concentration of power, the report argues, has also enabled lobbying influence that may have weakened UK regulation and competition.

Hardware dependencies present a further vulnerability.

The planned Loughton supercomputer is expected to rely heavily on chips from Nvidia, while the wider semiconductor supply chain is dominated by the US, Taiwan and South Korea.

Researchers estimate that 48% of facilities supplying the UK’s semiconductor sector are based in the US.

This leaves the UK exposed in the event of geopolitical disruption, with limited domestic capacity to sustain AI development independently.

While full technological sovereignty remains unlikely, analysts say the UK must prioritise diversification – combining domestic capability with a broader range of partners to ensure resilience.

As geopolitical tensions reshape global supply chains, continued reliance on US infrastructure risks becoming a strategic liability.

Without greater control over the systems that underpin AI development, the Sovereign AI Fund may strengthen innovation – but will fall short of securing sovereignty.

In the end, national security will depend less on sovereignty than on how the UK manages its reliance on others – and how far they can be trusted.

Stay ahead of emerging risk. Get notified whenever Sibylline publishes a new report.