Nvidia, the world's most valuable company, is donating $250m (£187m) worth of computing power to the UK to support research into the safety of artificial intelligence.

The US chipmaker, valued at about $5.5tn, said the pledge would run over five years and "support agentic safety and security research across the UK".

The announcement comes as concern rises about AI agents – systems that can carry out tasks on their own, such as browsing the web, writing code or making purchases, with limited human supervision.

Who gets the money

Nvidia said most of the funding would "support the wider UK AI safety and security research community". A portion will go to specific institutions, including the British Open-Ended Learning and Discovery Lab and the Science of Fundamental AI Research Lab, which is run by the universities of Edinburgh, Cambridge, Oxford and University College London.

Imperial College London will also receive support.

The donation takes the form of access to Nvidia's computing hardware rather than cash. That matters: the powerful chips needed to train and test advanced AI systems are expensive and scarce, and university researchers often struggle to get enough of them.

'World-class scientific rigour'

Jensen Huang, Nvidia's chief executive, said: "The UK brings world-class scientific rigour to this essential engineering challenge. With this $250m commitment, Nvidia will put accelerated computing in the hands of UK researchers to develop, test and strengthen safeguards for increasingly powerful AI."

Kanishka Narayan, the UK's AI minister, called the announcement a "vote of confidence in our world-leading strength and expertise in AI research".

The rise of AI agents

AI agents are the industry's latest focus. Rather than simply answering questions, they are designed to act – booking travel, managing email, running software and carrying out multi-step jobs on behalf of users.

That autonomy brings new risks. Safety researchers worry about agents being tricked into harmful actions, leaking data, or behaving in ways their designers did not intend. This week the AI company Anthropic warned that some AI models may be only months away from being able to copy themselves on to everyday smart devices.

Testing how such systems behave, and building defences against misuse, requires large amounts of computing power – exactly what Nvidia is now providing.

Edinburgh's role

The involvement of the University of Edinburgh will be welcomed in Scotland, which has a long history in AI research. Edinburgh's school of informatics is one of the oldest centres for the study of artificial intelligence in Europe.

Scottish researchers have increasingly focused on how to make AI systems reliable and trustworthy, and the new computing resources could strengthen that work.

Questions about influence

Not everyone will see the donation as purely altruistic. Nvidia's chips power most of the world's AI systems, and the company has a strong commercial interest in the industry continuing to grow and in governments seeing AI as safe enough to adopt.

Critics of the technology sector argue that safety research should be funded independently, so that it is not shaped by the companies whose products are being examined.

Supporters say the scale of the challenge means that public money alone is not enough, and that the industry has a responsibility to contribute.

Britain's AI ambitions

The UK has sought to position itself as a global leader in AI safety. It hosted the first international AI Safety Summit at Bletchley Park in 2023 and set up a government body to test advanced models.

Nvidia did not refer to that institute in its statement. But for the government, a $250m commitment from the company at the centre of the AI boom is a welcome endorsement of its strategy – and a reminder of how much British research now depends on American hardware.