UK Commits £40 Million to New AI Lab Targeting Hallucinations

The UK government announced on March 5 that it will invest up to £40 million over six years to establish a new Fundamental AI Research Laboratory. The lab, backed by the Department for Science, Innovation and Technology (DSIT) and UK Research and Innovation (UKRI), will also receive substantial in-kind access to AI Research Resource compute capacity worth tens of millions of pounds — a recognition that fundamental breakthroughs require not just funding but raw computing power.

Rethinking AI Architecture, Not Just Scaling Models

Unlike the dominant industry approach of throwing larger datasets and more compute at existing architectures, the new lab is explicitly tasked with rethinking how AI systems are built from the ground up. Its research targets include some of AI's most persistent technical failures: hallucinations, limited working memory, and unpredictable reasoning chains.

"This is a long-term investment in the brilliant minds who will keep the UK in the AI fast lane," said AI Minister Kanishka Narayan. The government expects the research to yield AI systems that are more accurate, transparent, and trustworthy — with downstream applications in earlier medical diagnoses, resilient infrastructure, and faster scientific discovery.

Applications Open Now Under DeepMind VP's Review

The funding call is open immediately through UKRI, with proposals assessed by a peer review panel chaired by Raia Hadsell, VP of Research at Google DeepMind and a DSIT AI ambassador. "The UK has the world-class talent and academic ecosystem to drive transformational research," Hadsell said. Eligible researchers are encouraged to bring their "boldest and most ambitious proposals."

The lab is the latest piece of UKRI's broader £1.6 billion AI strategy over four years — its largest single investment area for 2026-2030. That umbrella already includes the £27 million AI Safety Institute Alignment Project and record investment in compute infrastructure.

A Rare Bet on Fundamental Research

For AI researchers worldwide, the lab represents a rare publicly-funded commitment to fundamental — not applied — AI research at a time when most government funding gravitates toward deployment and regulation. The six-year horizon and dedicated compute allocation make it one of the more substantive non-corporate AI research opportunities currently available. Dr. Kedar Pandya of EPSRC said the initiative will "unlock fresh capabilities" by backing ambitious, ground-breaking work.

Researchers exploring how these UK opportunities compare to U.S. federal AI funding can find detailed analysis on the Granted blog.