Nvidia will provide the United Kingdom with $250 million worth of computing power over five years for research into the safety and security of increasingly autonomous AI systems. The commitment is valued at $50 million a year and will be delivered through UK-based Nvidia cloud partners rather than as a direct cash payment.
Most of the capacity will be available to the wider British AI safety and security research community. A portion will support the British Open-Ended Learning and Discovery Lab, the Science of Fundamental AI Research Lab—run by researchers at Edinburgh, Cambridge, Oxford and University College London—and Imperial College London.
The announcement came two days after AI minister Kanishka Narayan said scientists at the UK’s AI Security Institute were evaluating the latest generation of advanced models following work with leading developers and the US government.
The focus on autonomous systems reflects a growing security concern. AI agents can plan and execute sequences of tasks with limited human direction, creating the possibility of cyberattacks that spread or adapt faster than conventional software threats. Nvidia chief executive Jensen Huang said the donated capacity would allow UK researchers to develop, test and strengthen safeguards, while Narayan described the pledge as a vote of confidence in British AI expertise.
The company said distributed networks of inexpensive devices could eventually provide enough computing power for a model to reproduce itself without relying on a central data centre, although its tests also found that smaller models still made mistakes when combining the necessary steps end to end.
The UK government is preparing additional work on serious incidents involving autonomous AI agents and is considering stronger protections under its cyber-security and resilience framework. The new computing commitment gives researchers more infrastructure as the country combines pre-release model evaluations, public-sector cyber pilots and planning for potential AI-enabled attacks.