Government to set up Biosecurity Lab for biological AI threats
The government will invest £100 million in an AI Biosecurity Lab (AIBL) to help prevent biological risks caused by AI.
Anthropic and OpenAI will provide £200 million in computing resources and AI model credits.
As AI systems become more capable, they can create new biological risks. The AIBL will develop technical tools and skills to help the UK better understand, prevent, detect and respond to biological threats.
The lab will use advanced AI to support partners, including the UK’s national biosecurity centres and life sciences industry. It will help them spot biological threats earlier and speed up the discovery, development and testing of vaccines, treatments and other medical tools to fight threats. This will help the UK and other countries respond faster to disease outbreaks and better prepare for future biological risks.
The lab will bring together experts in advanced AI and biological science to develop practical technical tools. Its first work will focus on creating and using tools that help detect new biological threats early and speed up the development of vaccines, treatments and other medical tools to fight threats.
The lab will work with public-sector organisations, universities and other academic institutions, industry partners, and leading AI companies in the UK and other countries.
The government plans for the AIBL to receive funding from both public sources and charitable donors.
OpenAI will provide £100 million in computing resources and AI model credits. It will also give the lab trusted access to its most capable relevant AI models through the Rosalind Biodefense programme.
Anthropic will provide up to £100 million in AI model credits, trusted access to its models and dedicated technical support.
The AIBL will also work with Google DeepMind and Basecamp Research, a UK company developing new AI tools for life sciences, on research projects and AI tools for scientific work.
Basecamp Research will give the lab access to its AI models for biology, related tools and technical expertise to help address important biosecurity challenges.
Scientists and biosecurity experts from Google DeepMind will work with the AIBL on research projects. They will also explore giving the lab trusted access to advanced Gemini models and specialised AI models for science.
The AIBL will operate as an independent, non-profit organisation. Its goal is to make the UK a world leader in using AI to protect against biological threats. Like the AI Security Institute, it will attract leading technical experts from industry and universities. It will have enough independence to work quickly and focus on its goals, much like a company at the forefront of AI research and development.
The AIBL will receive advice from a scientific committee with experts from universities, industry and the third sector, which includes organisations outside government and business. More details about the committee will be announced later.
The AIBL is being launched as part of a National AI Biosecurity Initiative. This initiative brings together the UK Health Security Agency (UKHSA), the Animal and Plant Health Agency (APHA), the Defence Science and Technology Laboratory (Dstl), the AI Biosecurity Lab and the life sciences sector.
The AIBL will also work with the Medicines and Healthcare products Regulatory Agency (MHRA) to help prepare for health emergencies and guide the rules for medicines designed using AI.