UK Government Yet to Trial OpenAI Tech Months After Signing Partnership
A Freedom of Information (FoI) request has exposed that the UK government has not conducted any trials involving OpenAI's technology, eight months after signing a memorandum of understanding hailed as a partnership to harness artificial intelligence for addressing society's greatest challenges. The Department for Science, Innovation and Technology (DSIT) confirmed it holds no information on such trials, despite the agreement's promise to deploy advanced AI models throughout government and the private sector.
Lack of Progress and Transparency
In response to the FoI, DSIT pointed to an agreement allowing the Ministry of Justice to use ChatGPT with UK-based data storage, but this falls short of the broader ambitions outlined in the memorandum. Tarek Nseir, CEO of AI consultancy Valliance, which filed the request, criticized the situation, stating that rolling out ChatGPT in one department does not reflect the strategic intent of the partnership. He argued that this indicates either a failure in execution or intent, raising concerns about the government's seriousness regarding AI's economic impact.
DSIT highlighted ongoing work with the UK AI Safety Institute to test AI models and develop safeguards in collaboration with OpenAI, as well as initiatives with Nvidia and Nscale to deploy GPUs for Stargate UK, aimed at strengthening the UK's AI capabilities. However, apart from the MoJ's ChatGPT usage, none of these efforts appear to constitute the widespread deployment of advanced AI models as initially described.
Broader Concerns and Investigations
OpenAI responded that the FoI did not capture the full scale of its UK activities, expressing pride in the progress made. Meanwhile, a Guardian investigation found that Nscale's project to build the UK's largest supercomputer by 2026 is likely delayed, with the company misrepresenting its progress. Nscale's collaboration with OpenAI on Stargate UK, which aims to deploy 8,000 Nvidia chips, remains noncommittal, and OpenAI had nothing to share on its deployment timeline.
Matt Davies from the Ada Lovelace Institute emphasized that while AI could transform public services, government experimentation must be open and transparent. He noted that voluntary partnerships with big AI companies bypass usual procurement rules, raising accountability issues. Polling by the institute shows 84% of the public are concerned about the government prioritizing sector interests over protection, underscoring the need for a clear vision beyond merely increasing AI usage.
Additional Agreements and Future Plans
The UK government has also signed similar memoranda with Anthropic, Google DeepMind, and Nvidia. The Google agreement, concluded in December, is in early planning stages. Anthropic is developing an AI assistant for government services and collaborating on safety research with the UK AI Safety Institute. Nvidia did not comment on the matter.
This situation highlights ongoing challenges in translating high-profile AI partnerships into tangible public benefits, with calls for greater transparency and strategic focus to ensure AI genuinely improves lives.
