The UK government has signed a memorandum of understanding (MoU) with OpenAI in its latest push to attract AI investment and roll-out AI tools across public services.
OpenAI is expected to expand its London office as part of the move and explore where it could make investments in the AI Growth Zones set out in the government’s Compute Strategy.
The MoU also opens the possibility for OpenAI to build on its current relationship with the UK AI Security Institute, which investigates the security implications of AI, to establish a new program through which technical information could be shared or create new AI safeguards.
“DSIT and OpenAI share a vision for building on the UK’s strengths in science, innovation and talent, to maintain a world-leading UK AI ecosystem rooted in democratic values,” the MoU reads.
The government already uses the OpenAI model GPT-4o to underpin ‘Consult’, an AI tool for summarizing the thousands of comments received as part of public consultations.
“AI is a core technology for nation building that will transform economies and deliver growth,” said Sam Altman, CEO at OpenAI.
“Britain has a strong legacy of scientific leadership and its government was one of the first to recognise the potential of AI through its AI Opportunities Action Plan. Now, it’s time to deliver on the plan’s goals by turning ambition to action and delivering prosperity for all.”
As is standard for an MoU, the agreement is not legally binding and entered into by OpenAI voluntarily.
UK government continues to court AI developers
The MoU with OpenAI marks the third major big tech partnership the UK government has announced since the start of the year.
In February, the government announced an MoU with Anthropic, with an eye to improving public services, boost UK startups, and explore AI risks and opportunities.
Similarly, in early July the government announced a new partnership with Google Cloud to rid the public sector of “ball and chain” legacy tech.
This built on the government’s existing agreement with Google over ‘Extract’, an AI assistant for digitizing planning documents and increasing the speed of the planning process, which runs on Google Gemini.
Despite each MoU opening the possibility of each company investing further in the UK’s AI infrastructure, none have yet followed through. Critics will point to this, as well as the entirely voluntary nature of the agreements, to undermine their significance.
The announcements have come as Peter Kyle, the secretary of state for science, innovation, and technology, continues his year-long charm offensive with leading AI developers.
In the past year alone, Kyle has met with the executives of tech firms on several noteworthy occasions. It’s a campaign that has been criticized by the trade body UKAI for focusing too strongly on big tech.
At Google Cloud Summit London 2025, Kyle acknowledged this criticism, stating that he “plead guilty” to the crime of meeting with the sector more than his predecessor, Michelle Donelan.
“In truth, that was 28 times over the course of a six-month period,” he said. “That equates to about twice a week, over that time. Now, as technology secretary, I simply will not apologize for meeting the technology companies: that is my job."
Since being elected to office, the government has registered considerable interest in AI as a transformative technology.
The government has deployed its own AI tools for the civil service, such as ‘Humphrey’, a suite of AI tools intended to reduce public sector reliance on contractors and improve internal efficiency.
These solutions have been adopted in recognition of the need for trusted AI tools for internal use, as the data privacy risks associated with putting sensitive data into public AI tools becomes clear and organizations continue to tackle shadow AI - a trend in which workers use AI without oversight.
At the Google event, Kyle promised that “whenever possible, UK technology companies – large and small – [will] get a fair shot at winning a contract”. But to date, government AI rollout has been dependent on companies headquartered in the US.
Existing tools within Humphrey include the previously mentioned Consult as well as Redbox, a tool for tasks such as summarizing government documents, redrafting text, or expanding notes into longer copy.
The tool runs on OpenAI’s GPT-4o, Google’s Gemini 2, and Anthropic’s Claude 3, with civil servants able to choose which model they want to use at any given time.
In its algorithmic transparency record for Redbox, the government compares the tool to ChatGPT and Microsoft Copilot but notes that Redbox can be used for documents classified as ‘Official’ or ‘Official Sensitive’. This includes the bulk of government data as well as some information deemed especially sensitive.
The same document implies Redbox cannot be used for Secret or Top Secret information.
Make sure to follow ITPro on Google News to keep tabs on all our latest news, analysis, and reviews.
MORE FROM ITPRO
-
Why new hires are your weakest link when it comes to phishing attacksNews Research from Keepnet shows new hires are far more likely to fall for phishing attacks – here's how you can improve security awareness during onboarding processes.
-
The Channel Recruiter and Nebula debut new global talent resourcing solutionNews Tech Talent on Tap aims to provide channel firms with on-demand access to elite technical professionals from around the world
