TLDR : OpenAI and the UK government have agreed to collaborate on AI infrastructure, public service deployment, and safety, aiming for significant economic growth and enhanced national productivity.
The Department for Science, Innovation and Technology (DSIT) and OpenAI officially signed a Memorandum of Understanding (MoU) on July 21. This non-binding agreement will focus on three main areas: the development of AI-dedicated infrastructure in the UK, the deployment of OpenAI's cutting-edge models in public services, and the strengthening of their collaboration on AI safety.
This statement of intent, signed by OpenAI CEO Sam Altman and the UK Secretary of State for Technology, Peter Kyle, will contribute to achieving the goals of the AI Opportunities Action Plan announced in January 2025 by the Labour government to revive the country's economy. It aims to make AI a growth driver, with estimated gains of £47 billion per year and a 1.5% increase in national productivity.
As part of the protocol, OpenAI could invest in AI Growth Zones, regions identified by the government to host strategic infrastructure such as data centers and research hubs. Funded with £2 billion by the government, this program has attracted over 200 applications across the UK, including Scotland and Wales. The goal is to attract additional investments in these territories while strengthening cooperation between universities, startups, and public institutions.
The UK government is already leveraging OpenAI's technology to increase the efficiency of its public services. The GPT-4o model powers several modules of the AI assistant Humphrey, deployed within the civil service. Among them, the Consult tool automates the processing of responses during public consultations, significantly reducing the time required for analysis. The agreement paves the way for the gradual integration of OpenAI's models into strategic areas such as justice, education, defense, or homeland security.
Technical exchanges with the British AI Safety Institute will also be strengthened. OpenAI will provide information on the capabilities of its models and the potential risks associated. This sharing aims to improve government understanding of the challenges related to advanced AI systems while informing discussions on necessary standards and safeguards. The two partners will co-develop research programs to anticipate and mitigate potential risks.