London launches an AI lab against biological threats, funded largely by OpenAI and Anthropic credits

Three hundred million pounds: that is the figure the British government highlights in its October 24 press release. It announces the creation of the AI Biosecurity Lab (AIBL), a lab tasked with using the most advanced artificial intelligence to detect biological threats earlier and speed up the development of vaccines and treatments. Bloomberg had revealed the project the day before.

But the details matter. The British state is contributing £100 million. The rest, up to £200 million, is pledged by two American companies, OpenAI and Anthropic, and comes mostly in the form of credits to use their models.

What each company is really contributing

According to the official press release, OpenAI is committing £100 million in compute and model credits, plus privileged access to its best-performing biology models, as part of its Rosalind Biodefense program. Anthropic is pledging up to £100 million in model credits, privileged access to its models and dedicated technical support.

Google DeepMind will also take part, but without a figure: its researchers will collaborate on projects, and the company is "considering" opening access to its Gemini models and to specialized scientific models. The British company Basecamp Research will contribute its AI models applied to biology.

A lab with no face yet

The AIBL will be an independent nonprofit organization. Like the UK's AI Security Institute, which tests AI models before their release, it will need to attract technical talent from industry and research. It will work with the public agencies for health, animal health and defense research (UKHSA, APHA, Dstl), as well as with the medicines regulator, the MHRA, on the regulation of AI-designed drugs.

Several points remain unclear. The press release names no leader and gives neither a location nor a start date. Details on the scientific advisory board will be provided "in due course." The text also does not say for how long the credits will be available.

What this says about the British strategy

The setup raises a question of dependence. A large share of the lab's resources will be consumed on the platforms of the two companies that fund it, even as those companies develop the very models whose biological risks worry authorities.

For a European executive, the announcement also shows how London funds its AI policy: a limited public stake, complemented by partnerships with the major American vendors. The names of the leadership and of the scientific board, which the press release does not yet give, will show who will really run the lab.