The UK government has kicked off a £500m initiative to aid domestic AI startups. Launched by Liz Kendall, Secretary of State for Science, Innovation and Technology, the new Sovereign AI Unit is being pitched as an essential pillar of the UK’s tech industrial strategy.
The fund, chaired by Balderton partner James Wise, is being positioned as a government-backed vehicle to invest in strategically important UK AI companies and support them with capital, compute and access to government.
The state-backed investment fund will target British AI firms, hoping to keep them in the UK while scaling internationally.
A statement said the Unit will act “as a venture capital fund within government, making targeted investments in UK AI firms.”
The Sovereign AI Unit will also provide access to national compute infrastructure like supercomputers, research support, procurement opportunities and regulatory guidance.
The UK claims it has more than 5,800 AI tech companies and 200+ unicorns, laying claim to the largest AI ecosystem in Europe. The launch of the unit comes a day after London-based autonomous driving firm Wayve secured an additional $60m investment AMD, Arm and Qualcomm.
The first cohort of companies receiving backing from the fund is expected to be announced alongside the launch.
The launch of the Sovereign AI Fund has been broadly welcomed by investors and startup backers as a necessary signal that Britain intends to compete in the global AI race.
Messaged a number of Investors and founders to gauge their reaction to the launch.
Aside from over positivity, the reaction from the venture community is not unqualified enthusiasm. The dominant view is that more capital, compute access, and government engagement are good for UK AI companies, but if the Fund tips over into direct state investment, that could carry real risks.
Investors counselled that the Unit should aspire to commercial discipline, political independence and a clear mandate to work closely with private capital rather than supplant it.
For many investors, the launch is a positive strategic signal. Tom Wilson, partner at Seedcamp, called it “another positive addition to the already booming UK AI ecosystem,” saying the UK already has “some of the best talent working in AI” and that further “capital, compute, and government access is great news for founders.” Pippa Lamb, VC and angel investor at Sweet Capital, was similarly upbeat: “It’s time for the UK to be more ambitious…I’m particularly encouraged to see that SovAI is taking practical steps… providing the companies it backs with compute.”
That enthusiasm was echoed by Ziv Reichert, investor at Phoenix Court, and the home of VC LocalGlobe, who argued that the headline number matters less than the direction of travel. “This is a strong step towards building deeper support for UK AI capabilities,” he said. “The UK already has exceptional AI talent, but the fund could help promising AI and deeptech ventures “move faster, go further and, most importantly, retain talent in the UK for longer.” He was joined by Hussein Kanji, partner at Hoxton Ventures, who said “it makes sense for government to play a more active role, not just to be a regulator or consume the technologies, but to participate in the wealth creation.”
Others saw the move as geopolitically significant. Viktorija Trimbel of Coinvest Capital said Europe is “lagging behind in multiple critical industries” and needs “large volumes of capital to catch up quickly, to build resilience and sovereignty.” The UK acting post-Brexit, she argued, “puts implicit pressure on EU member states and the Commission.” It also creates “both a competitive and a collaborative dynamic simultaneously, and ultimately one that should make Europe more resilient and secure.”
But the sharpest concern is whether the government should invest directly in companies at all. As Trimble characterised it: “political interference in deal selection, slow deployment cycles, and risk aversion dressed up as due diligence.”
Several other investors argued that the state is better suited to acting as an anchor LP in specialist funds, or as a customer, rather than trying to become a VC itself.
Sarah Antor of Helena Capital said she supported public investment programmes in principle, and has personal experience of this, but warned of “the high risk of incompetence and corruption.” Her view was blunt: “I think the UK government is being a bit short-sighted with the idea of direct investments. There is a massive need for capital, but the biggest need is growth capital, and the government is not equipped to turn itself into either an early or late stage VC overnight. It makes much more sense to award those pounds to already established UK or European VCs.”
Richard Muirhead of Fabric Ventures, while praising Wise’s appointment, made a similar point. “James Wise is exactly the right kind of chair for this, a founder-facing investor who understands what scaling companies need,” he said. But he warned that “direct government investment into companies consistently underperforms private VC,” while government-backed fund-of-funds models tend to perform better when professional managers have “commercial skin in the game.” His recommendation: deploy the £500 million “through established UK specialist fund managers who’ve been backing AI and deep tech without domestic institutional support.”
That view was not universal. Daniel Dippold from EWOR VC Fund said he “loved” the move, even while doubting that the government would immediately be good at picking winners. But he added that the long-term upside could be government learning how power-law venture outcomes shape the economy. “More capital lifts the tails of the distribution curve and almost always comes with higher eventual socioeconomic output.”
Kinga Stanislawska, Founder, Experior VC (and Founder, European Women in VC), drew on her experience with the European Innovation Council Fund. “This is a good start and an important strategic signal for the UK, even if £500 million remains modest relative to the scale of capital now flowing into AI globally,” she said. “The real test is execution. Direct government investment works best when it is catalytic: co-investing with strong private investors, supporting capability gaps, and strengthening the ecosystem without distorting valuations or imposing terms that make later fundraising or exits harder.” This was echoed by Pedro Santos Vieira, Partner, 500 Global who said the move “reflects a broader European realization that AI is no longer just a venture theme; it is an industrial policy theme.”
A number of investors welcomed the fund only if it brings in more private capital, rather than crowding it out. As Trimbel put it: “The real question isn’t whether £500 million is enough (it isn’t) but whether the fund is structured to catalyse private capital alongside it. A sovereign fund that crowds in private VC is a very different animal from one that crowds it out.”
There is also a question of whether the UK should be investing in startups, or buying from them. Andrew J Scott of 7percent Ventures said “£500m for UK AI is no bad thing,” but asked whether government should be “investing into, instead of buying from, UK startups.” His alternative was more direct: “I’d like to see a £500m initial contract for a UK sovereign LLM.”
Amelia Armour of Amadeus Capital Partners echoed that sentiment, arguing that the fund could “speed up engagement between UK AI firms and government departments.” In her experience, “having the UK govt as a customer provides great validation when selling to others.”
The sceptics were not dismissive of the ambition, but largely of the mechanism. One anonymous European VC said that while “winning in AI matters a lot and every initiative counts,” the fund may be too small to move the market if UK AI startups are already on a roll.“ Considering UK AI start-ups raised $5.8b in q1 2026 alone, how big a difference does a £500m fund investing even £25m a quarter makes?” they asked, warning that public funds investing directly have “a poor track record” and suggesting an “ARIA-like approach” may have been more effective.
Ekaterina Almasque, partner at BlankPage Capital, also welcomed the initiative, but was concerned the “the goals and the mandate” were “largely unclear” and that the UK AI sector was already seeing “record investment”.
The overall sentiment, then, is cautiously positive, especially towards both the new fund and the access to more compute, but a vehicle that might ‘pick’ companies could distort the market.
UK founders welcome SovAI, but warn capital alone won’t fix scale-up issues
The Sovereign AI Fund was broadly welcomed by founders as a serious signal that the UK is ready to back its own AI ecosystem. But they also raised concerns about speed, procurement, compute, risk appetite and access to real deployment pathways.
Founders were encouraged by the signal that the government sees AI as national infrastructure, but are worried about the initiative being slow, political, bureaucratic or skewed towards the most obvious, capital-intensive companies, over faster-moving AI companies.
Alexander Zheltov, founder of Provos and Flylane, said the fund could help create a new kind of UK founder base, but “execution matters more than intent.”
Several founders argued that the real value of the fund may not be the £500 million itself, but what comes with it. Will Taylor, founder of writing tool startup Versey, said: “The real value here isn’t the £500m, as there is plenty of capital available, it’s the compute access and procurement commitments.” For early-stage founders, he argued, a government-based first customer could matter more than just another cheque. “If the government becomes a genuine first buyer through the Advance Market Commitments, that changes the maths for early-stage AI companies far more than another cheque would,” he said.
That point was echoed by Bridget Harris from YouCanBookMe, who said the move was “good news overall” but asked, “who will be making the investment decisions… for every 1 person that says yes, there’s 9 people who love to say ‘no’ in govt — there is very little risk appetite.”
Others warned that the UK’s AI challenge is not only financial. Catherine Jeong, co-founder of NODI Energy, said the move was “directionally right,” but argued that the real constraint is whether AI can scale physically. “The UK has AI ambition but on a Victorian grid,” she said. “The constraint now is execution: build times and grid connection timelines.”
Andy Cairns, founder of RAVL, said the fund could be “a genuinely positive move” if it backs more than just frontier models and centralised infrastructure. His concern is that government capital may gravitate to the ‘loudest’ and most capital-intensive parts of the AI stack, while overlooking practical approaches. “If the Sovereign AI Fund is willing to support this kind of approach, not just frontier models but applied, privacy-first, energy-efficient systems, it could unlock a very different layer of innovation in the UK,” he said.
The strongest structural critique came from founders looking at Europe’s long-running scale-up problem. Laura Towart, founder and CEO of Vivan Therapeutics, said direct government investment is not without precedent, pointing to the Covid-era Future Fund. But she said the crucial question is “if this fund is designed with genuine risk appetite.” Ken Donald, Co-founder and Co-MD, OPUS added that there should be “equivalent focus dedicated to encouraging and supporting more people into entrepreneurship for the first time.”
Markus Törstedt of Akka VC was more sceptical about public capital as a substitute for private institutional investment, which is lacking in Europe. Government-backed funds and startups do not always optimise for “shareholder and tech ecosystem value creation,” and can create distorted incentives for both investors and founders, he said.
It seems that both founders and investors, while welcoming the initiative (especially around investment and access to compute) are clear-eyed about how it might go wrong.
