AI Biosecurity Lab gets £100m, but £200m of its backing is compute credits

The UK government will establish a national AI Biosecurity Lab with £100 million in public funding, announced as part of Wednesday's Budget. Officials describe it as a "world first." AI Minister Kanishka Narayan called it "a major new global institution" bringing together "world-leading scientific expertise."

The lab is not running. It is still being set up, and the full details of its role and remit are yet to emerge.

The larger number in the announcement is not public money. OpenAI and Anthropic are providing up to £200 million in combined support, which the BBC reports comes as AI model credits rather than cash. Google DeepMind and Basecamp Research are named as research partners.

That structure is the story. A state-backed biosecurity body is being stood up to help govern AI-enabled biological risks, and a substantial share of its backing is in-kind compute from two of the frontier AI firms whose models sit inside that risk category.

The £100 million is a new budget allocation, not a previously announced commitment.

What the lab will actually do remains open.

The announcement converts a long-running policy concern into a funded institution, which is a real step. Whether it delivers practical defenses depends on decisions not yet made: what the lab is empowered to do, how compute contributions are counted and governed, and whether a body partly resourced by the firms it may need to scrutinize can hold its independence.

For now, the verifiable position is narrow. Britain has committed £100 million and named partners. The lab is being built. The rest is pending.