UK invests £60M in two new AI labs at Oxford and UCL to build cheaper, open-source AI

The UK is betting that the future of AI doesn't have to be bigger, more expensive, or locked inside the data centres of a few American tech giants. On 23 June 2026, the Department for Science, Innovation and Technology (DSIT), through UK Research and Innovation (UKRI), announced up to £60 million in funding for two new AI research labs at the University of Oxford and University College London (UCL) .

The two labs: SOFAIR and BOLD

Each lab has a specific, complementary mission, backed by funding from the Engineering and Physical Sciences Research Council (EPSRC) and access to large-scale computing power .

SOFAIR Lab (Science of Fundamental AI Research) — led by UCL

Led by Professor David Barber at UCL, in collaboration with the universities of Cambridge, Oxford and Edinburgh, SOFAIR's mission is to develop new open-source AI technologies that can run on widely available hardware, including ordinary consumer computers . The lab brings together researchers from computer science, mathematics, statistics and neuroscience to design fundamentally new AI architectures, directly addressing current problems like AI systems giving inaccurate responses .

BOLD Lab (British Open-ended Learning and Discovery) — led by Oxford

Led by Associate Professor Jakob Foerster at Oxford, with UCL and other partners, BOLD aims to rethink how AI learns from the world, pioneering new approaches that enable AI to learn without requiring vast centralised computing power . Instead of simply scaling up existing methods, it will explore new paradigms for learning, aiming to build AI that is more efficient, more open and better aligned with human needs .

The strategy: efficiency over scale

The entire initiative is a deliberate alternative to the "bigger is better" approach of large US tech firms . The strategy isn't to outspend the biggest technology companies on data and compute, but to out-innovate them on efficiency, openness and accessibility .

• Lower costs through fundamental efficiency: Instead of trying to compete on scale, the labs target fundamental changes to how AI works, making it cheaper to run and accessible to far more organisations .
• Open-source, hardware-agnostic AI: SOFAIR will build open-source models that run on widely available hardware, reducing the need for expensive, specialised infrastructure .
• Reducing dependency on centralised data centres: BOLD's work on decentralised learning aims to reduce dependence on remote data-centre infrastructure owned by big providers .
• UK sovereign capability: AI Minister Kanishka Narayan stated the initiative strengthens UK expertise "at home" and "reduces reliance on others" . Professor David Barber explicitly said SOFAIR will "reduce dependency on the small number of model providers, boosting UK sovereignty" .
• Access for smaller businesses and public services: By making AI cheaper and easier to adopt, the labs aim to open AI to small businesses, public services and researchers across the UK who currently cannot afford top-tier AI systems .

In short, the UK is pursuing a path of sovereign, open, and efficient AI, with the potential to reshape how AI is built and deployed — not by building the biggest model, but by building one that almost anyone can run.