
Britain Just Bet Half a Billion Pounds on Keeping Its AI Startups. The Government Is Acting Like a VC.
The UK launched a £500M Sovereign AI Fund and named its first batch of startups. The pitch: start, scale, and stay in Britain.
The AI Post newsroom — delivering AI news at the speed of intelligence.
The UK government just launched a £500 million fund to invest in British AI startups, took equity stakes in its first two companies, and gave six more access to government-funded supercomputers. The fund is called the Sovereign AI Unit. It is designed to operate like a venture capital firm, not a government grants office. That distinction matters.
Technology Secretary Liz Kendall announced the first investments on Thursday from the London offices of Wayve, the self-driving car startup now worth $8.6 billion. Her message: "We have to seize this to make it work, for Britain, for our jobs, for solving the biggest challenges we face as a world."
The fund's first direct investment is an undisclosed shareholding in Callosum, a London startup that helps different types of computer chips work together to train and run AI models more efficiently. A second, unnamed company also received an equity stake. Six additional startups will get access to a network of government supercomputers in exchange for a "right of first refusal" to invest.
The Startups
The first cohort includes Prima Mente, building "biological foundation models" to tackle diseases like Alzheimer's. Cursive, founded by Google DeepMind alumni, is developing autonomous AI agents. Odyssey builds "world models" where AI systems interact with convincing simulations of the real world. Each company received one million GPU-hours of compute access.
The compute-for-equity model is the interesting part. The government provides supercomputer access. In return, it gets first refusal on investment rounds. If these startups take off, the UK taxpayer holds real equity in real companies. If they don't, the cost is compute time that would have sat idle anyway.
The Brain Drain Problem
This is not altruism. The UK has a brain drain crisis. British AI founders consistently raise seed funding at home and then move to the US for Series A and beyond, because American VCs write bigger checks. Stanford's 2026 AI Index documented an 89% drop in AI talent migrating to the US, but the UK talent leaving for Silicon Valley has been a structural problem for a decade.
Chancellor Rachel Reeves was blunt: the goal is ensuring that internationally competitive companies "start, scale and stay here in Britain." The timing is not accidental. Anthropic just committed to an 800-seat London office. OpenAI secured 544 seats days before that. Both companies are hiring aggressively from British universities and AI labs. The Sovereign AI Fund is the government's attempt to compete with those checkbooks.
The Risk Nobody Is Talking About
Kendall acknowledged "people are worried about the risks and what it means for their jobs." In January she admitted that "some jobs will go" as AI automates certain tasks. But she argued that AI entrepreneurs believe they can "create jobs" too.
That is the uncomfortable bet at the heart of every government AI strategy: spend public money to build the technology that might eliminate public jobs. The UK is making that bet with half a billion pounds and a VC structure that at least gives taxpayers a share of the upside. Most governments are not even getting that much.
Danyal Akarca, co-founder of Callosum, called the UK the "natural place" to build his company, citing university talent and private AI labs like DeepMind. Whether that remains true depends on whether £500 million is enough to outbid Silicon Valley. History suggests it is not. But at least Britain is finally in the game.
First reported by The Guardian. Additional reporting from Gov.uk, Sifted, and Tech Funding News.