Britain's Cabinet Office announced on Saturday that the government will invest £100m in a new AI Biosecurity Lab designed to leverage cutting-edge artificial intelligence systems for identifying and responding to biological hazards. OpenAI and Anthropic will together contribute up to £200m in the form of computing credits and model access, rather than direct funding.
The government has designated the facility as the world's first dedicated AI biosecurity organisation. Its initial focus will centre on two priorities: early detection of emerging biological threats and acceleration of vaccine and treatment development efforts.
OpenAI's contribution comprises £100m in computational resources and model credits, alongside access to its systems through the Rosalind Biodefense programme. Anthropic will supply up to £100m in model credits, model access, and technical assistance. Google DeepMind has committed to participating in research initiatives and may provide access to its Gemini and science models. UK-based Basecamp Research will contribute its AI biology models to the effort.
The lab will operate as an independent non-profit organisation structured similarly to the AI Security Institute. The announcement did not specify a director, location, or launch date, nor did it provide a timeline for deployment of the £100m government allocation.
Multiple government agencies will participate in the initiative, including the UK Health Security Agency and the Defence Science and Technology Laboratory. The MHRA, Britain's medicines regulator, will collaborate with the lab on regulatory frameworks for AI-designed pharmaceuticals.
This lab forms part of a broader £600m+ AI security investment package unveiled ahead of Wednesday's Budget announcement. The package also encompasses £120m allocated to the AI Security Institute. Chancellor John Healey unveiled the broader initiative, while AI minister Kanishka Narayan described the lab as a significant new global institution.
The announcement mirrors a recent commitment from Nvidia, which pledged $250m in computing power to support UK AI safety research earlier this month, similarly offering computational resources rather than monetary funds.
Source: The Next Web
