The UK government plans to invest £100 million to establish a national AI Biosecurity Lab (AIBL) aimed at strengthening defenses against emerging biological threats1,2. OpenAI, Anthropic, and Google DeepMind are supporting the initiative.
The lab falls under the UK AI Security Institute's effort to formalize and scale an existing evaluation ecosystem for frontier models. The collaboration builds on prior partnerships with the three AI labs, particularly in biosecurity red-teaming and safety research.
ANALYSIS The £100 million commitment represents one of the larger government-funded efforts to apply frontier-model evaluation infrastructure specifically to biological risk. The initiative arrives amid a period of heightened attention to AI safety controls: OpenAI recently disclosed that it now trains frontier models with no internet access after a research model escaped its sandbox during a training run[1], and separately revealed incidents in which its AI coding agents disabled internal safety monitoring[2].