Travers Smith's Venture Insights: Britain's Sovereign AI Bet

Whatever the strategy, the judgement and experience of your investment team is the key to success. And, in launching the sovereign AI fund – a £500m venture capital fund – the UK government seems to have got that right. Leading the charge is James Wise, partner at Balderton.

So what is it that they have to deliver?

The UK Labour government has been exploring ways to support growing companies both through enabling venture capital firms and through early-stage government investment. Labour policy appears to be influenced by the work of Mariana Mazzucato, who has argued (in The Entrepreneurial State) that government support to early-stage companies should form part of a public-private partnership. On this model, governments provide public funding to bridge the "Valley of Death" in early-stage companies through a combination of grants, loans and equity. Governments retain the equity stake when venture capital provides further investment, so the up front risk across a range of companies will be repaid by the growth in equity value for a few stars.

Sovereign AI is a government-backed fund, aiming to be as responsive and agile as any VC fund. Part of the UK's AI Opportunities Action Plan (which places AI at the heart of the UK's economic and national security ambitions), the Fund's mission is for founders to start in Britain, scale in Britain, and win globally.

That framing is important. The objective is not necessarily to scale companies to compete with the likes of Anthropic, but to build on the existing strengths within UK industry and ensure that the UK has a seat at the table in shaping the technology. Even so, can a fund of this scale deliver that goal?

The government is being creative in designing a fund that's not just a cheque. Backed companies get fast-tracked visas for R&D talent and access to the UK's largest supercomputers and funding to build datasets, infrastructure and tools that accelerate growth – this kind of compute is usually the preserve of the biggest tech companies. In exchange, the fund can take rights of first refusal when a company raises its next round. Also on offer is the payment of legal fees if any overseas startup wants to "flip" to a UK limited company as well as government procurement with the government acting as an early customer and serving as a quality indicator to other investors.

The investment strategy is relatively narrow (or deliberately focused – depending on your perspective). Ticket sizes are between £1m and £10m, investing directly into early-stage UK AI companies alongside co-investors on commercial terms – this is not a grant and the aim is that the UK taxpayer benefits in some of the upside.

The fund will work alongside BBB (the British Business Bank), but the two have different strategic focus and different ways of investing. BBB typically invests in investment funds, which in turn invest in companies (for a VC investment). The new sovereign AI fund invests directly in the UK AI companies. It picks companies – that is a different risk, but a bet the government will need to make. Whether the sovereign AI fund can operate with the pace that early-stage investing demands remains to be seen – government and speed are not natural bedfellows.

AI governance and security are a particular focus. This ties in with previous statements by the UK government around responsible AI. The speed at which capable AI models are advancing – Anthropic's latest model being a striking example of how quickly the frontier is moving – means that governments which are slow to act risk being left behind and strategically exposed.

"The UK's problem has not been a shortage of seed funding – it has been the growth stage where domestic capital runs thin"

But will this be enough?

The UK has struggled to keep its best founders at home. Many drift to the US, drawn by deeper pools of capital, larger markets and the gravitational pull of Silicon Valley. The fund does not require its investee companies to remain in the UK. It uses incentives, not mandates. That is probably the right approach – mandation rarely produces the best companies, but it does mean the fund's success depends on those incentives being genuinely attractive relative to what founders can get elsewhere.

£500m spread across early-stage investments of up to £10m is not transformative on its own and is a modest sum measured against the capital being deployed by other countries and large technology companies in AI globally. For example, the EU's Scaleup Europe Fund, while not yet in operation, is a €5bn fund, which, it has just been announced this week, will be managed by EQT. The fund will make direct equity investments of €100m+ in promising European companies in strategic deep tech areas. The intention is that it will lead investment rounds.

The UK government is placing a strategic bet. Whether this is enough to nurture future AI champions on British soil and keep them there remains to be seen. The UK's problem has not been a shortage of seed funding – it has been the growth stage where domestic capital runs thin and this fund does not obviously fix that.

That said, the government deserves credit for a strategy that acknowledges the UK's existing strengths, building an expert commercial team and proffering both financial and non-financial support. The fund and its ambition is a positive step. Its ability to attract private capital alongside government support – and whether UK startups are more likely to thrive as a result – will be the real measure of its success.