Nvidia announced on October 14 a five-year compute donation valued at $250 million to strengthen agentic AI safety and security research in the United Kingdom. The commitment provides $50 million in compute access a year, delivered through UK-based Nvidia cloud partners.

Most of the compute will go to the wider UK safety research community. A share will support three named efforts: Oxford-led BOLD, UCL's SOFAIR lab, and Imperial College London work on robotics safety and fundamental agentic models.

The pledge lands as Britain pushes to be a centre for AI evaluation through its AI Security Institute, and a day after OpenAI and Anthropic told UK MPs that safety duties should be binding in law.