The UK has inked a major new deal with Google DeepMind that aims to turn AI research into tangible benefits for the British workforce, deepening the country’s controversial ties to one of the world’s most powerful tech firms.

With the deal, the frontier AI firm will provide UK scientists “priority access” to some of the most advanced tools of the AI Age, with the government hoping any subsequent breakthroughs will translate into real-world gains across areas like defence, medicine, energy, and public services.  

To that end, DeepMind said it will open its first automated research lab in the UK next year, a facility to be built from the ground up to be fully integrated with the tech firm’s Gemini AI. Run by researchers, the automated lab will harness cutting-edge tech and robotics, and will initially focus on developing new superconductor materials that can carry electricity with no resistance. 

The government said that these new superconductors could help deliver low-cost medical imaging, ensure next-gen computer chips are more efficient, or even transform transport, accelerating the UK’s path to net zero.

The partnership will also see DeepMind work with the government and the UK’s scientific community on research in areas like fusion energy, the creation of smart public services, education, and responsible AI development.

That plan hinges on a new partnership with the UK AI Security Institute, sealed by a fresh MOU that aims to advance security and safety research for future AI models through deeper collaboration across joint research areas.

For its part, DeepMind said that its collaboration with the UK will “offer a blueprint for other countries” to translate AI innovation into public benefit, with the firm’s goal being to empower the next generation of scientists through its suite of AI tools. 

“This partnership will make sure we harness developments in AI for public good so that everyone feels the benefits,” said the prime minister, Keir Starmer.

“That means using AI to tackle everyday challenges like cutting energy bills thanks to cheaper, greener energy and making our public services more efficient so that taxpayers’ money is spent on what matters most to people.

“This is national renewal in action – driving innovation to make our country stronger and fairer for everyone.”

Keen to see emerging technologies like AI meaningfully contribute to economic growth, this latest move comes hot on the heels of several other government efforts to tap into Google’s extensive reach.

Earlier this summer, the government entered an agreement with Google Cloud to modernise outdated public sector IT and retrain thousands of civil servants with digital skills. 

Following that, ahead of President Trump’s state visit in September, the government announced that Google would open a new data centre in Hertfordshire as part of a two-year, £5 billion investment, coming as part of a package of similar investments from other major tech firms.

Recommended reading

• Report: UK IT Leaders ‘Flying Blind’ on Cloud Spend
• 60% of UK IT Leaders Urge for Data Sovereignty In Wake of US Tariffs
• Report: Cloud Complexity is Outpacing Cyber Defence

However, the public sector’s tightening alliance with companies like Google has already sparked tensions.

Research from the Ada Lovelace Institute this year, for example, found growing concerns about the role, motivations and impacts of private companies developing and using AI in the public sector, with almost a third of the general public being “very” worried about how businesses produce and use tech in welfare and public health.

The question of how, exactly, tech firms like DeepMind will access and use the troves of public sector data the government holds, which is essential to building efficient AI tools, has already been a contentious issue.

In 2015, the Royal Free London NHS Foundation Trust shared vast amounts of sensitive health information from 1.6 million patients with DeepMind to create a life-saving app, Streams, that could spot when patients were at risk of kidney injury. 

But, according to the ICO, that information was given without patients’ knowledge or consent, and, crucially, without the option to opt out, an issue which resulted in a class action lawsuit against Google and DeepMind, a claim later rejected by the High Court.