The UK government has unveiled a sweeping £600 million ($800 million) AI security package, its largest single commitment yet to defending against AI-enabled threats. The centerpiece is a new AI Biosecurity Lab, described as the world’s first dedicated AI biosecurity organization, backed by £100 million in government funding and up to £200 million in commitments from frontier AI companies.
OpenAI will provide £100 million in compute and model credits along with access to relevant models through its Rosalind Biodefense program, while Anthropic will contribute up to £100 million in model credits, trusted model access, and technical support. Google DeepMind and life-sciences startup Basecamp Research join as technology and research partners. An additional £65 million is earmarked for developing countermeasures against pathogens.
The lab will work with public-sector agencies to accelerate detection of emerging biological threats and speed up the development of vaccines and treatments. Beyond biosecurity, the package allocates £120 million to the existing AI Security Institute, £130 million for cyber defenses on critical infrastructure, £60 million for AI-driven robotics and manufacturing, and smaller amounts for a chip design academy, regional AI initiatives, and research into AI’s economic impact.
The announcement highlights a growing policy focus on frontier-model risk as leading AI models grow more capable — compare them in our ChatGPT vs Claude comparison. With open-weight models now widely distributed through Hugging Face and runnable locally with tools like Ollama, governments are racing to keep safety research ahead of open deployment.
Source: Business Today (via Reuters)
