According to Reuters, Nvidia has pledged to provide Britain with computing resources worth $250 million over five years. The funds are meant to support AI safety research in the country. According to Reuters, this amounts to accelerated computing power worth $50 million per year. It is to be used, among other things, for testing the safety of AI agents. Nvidia founder and CEO Jensen Huang said, according to Reuters: "AI capability and safety must advance together."
Background
According to Reuters, Britain is not home to any major AI developers but wants to play a leading role in evaluating and managing the risks of AI systems, partly through its AI Security Institute. The pledge comes amid growing scrutiny of AI safety after developers such as Anthropic and OpenAI disclosed instances of undesired behaviour by AI agents. Reuters also reports that Microsoft President Brad Smith recently said the UK was well placed to lead work on AI safety guidelines.
What the pledge covers
According to Reuters, the donation is intended to connect several British research areas, including BOLD at the University of Oxford, which develops approaches that help AI systems learn, adapt and operate alongside people, as well as robotics safety work at Imperial College and programmes at University College London.
Context
For German companies, the main point is that safety questions are increasingly being publicly funded and institutionalised by large providers. Whether this leads to practical testing standards that matter beyond Britain remains open and is still to be seen.
Sources
Editorially owned by Ideal Syka. Sources and method: Newsroom & method. Tips and corrections: ai@i6eal.de.
