Skip to main content

New Turing AI Fellows will help strengthen the UK's AI capability

Two University of Oxford academics are part of the new cohort of Turing AI Fellows, announced today by UK Research and Innovation (UKRI). Delivered by the Engineering and Physical Sciences Research Council (EPSRC) and backed by more than £39 million in government funding, the 20 new fellowships will help strengthen the UK's AI capability by both advancing core AI research and accelerating the adoption of AI across disciplines.

Abstract illustration of  Abstract program code, big data, artificial intelligence, machine learning, seen as computer code suspended among grid lines.

Two University of Oxford academics have been awarded prestigious Turing AI Fellowships to undertake creative and innovative AI research. Image credit: MF3d, Getty Images.

Professor Charlotte Deane, Executive Chair of EPSRC and a Professor at Oxford’s Department of Statistics, said: ‘AI has enormous potential to transform the way we understand and tackle some of the biggest challenges facing society. These fellowships will give outstanding researchers the freedom and support to push the boundaries of AI itself, while also applying it in new ways across a wide range of disciplines.’

About the new Oxford Turing AI Fellows:

Professor Jocelyn Monroe

Professor Jocelyn Monroe of Oxford University’s Department of Physics has been awarded a three-year Turing AI Pioneer Interdisciplinary Fellowship. She will use the search for dark matter to develop more trustworthy AI systems capable of identifying extremely rare events. 

Dark matter makes up most of the matter in the Universe, yet its fundamental nature remains one of science’s greatest mysteries. Although invisible, its gravitational influence is essential to our understanding of how galaxies form and evolve. Scientists are searching for the particles that make up dark matter using extraordinarily sensitive detectors, deep underground, designed to observe the tiny signals produced when dark matter particles interact with ordinary matter.

Finding these elusive particles is one of the most challenging searches in modern physics.  Using the Fellowship, Professor Monroe will build AI systems that can reason reliably about needle-in-a-haystack events, including recognising when reliable detection is not possible. 

As a testbed, she will use the international DarkSide-20k experiment, currently under construction, where genuine signals will be billions to trillions of times rarer than background noise. By stress-testing AI under these extreme conditions, the research will pioneer methods that are not only more sensitive, but also more reliable, interpretable, and honest about uncertainty. 

The resulting methods could have applications far beyond fundamental physics, including cybersecurity, critical infrastructure monitoring, and other areas where reliably identifying rare events — while avoiding false alarms and misplaced confidence — can make a profound difference.

Portrait photograph of Jocelyn Monroe, a white lady with brown hair
“This Fellowship comes at an incredibly exciting moment in my career. Having spent much of my research life pushing the boundaries of dark matter detection, I now have the opportunity to take on a new challenge: developing AI that can transform how we search for the rarest events in nature. The Fellowship gives me the time, resources and collaborations to build deep expertise in AI, bringing together two rapidly advancing fields in ways that would not otherwise be possible.”
— Professor Jocelyn Monroe, Department of Physics

Professor Denny Wu

Professor Denny Wu (who will relocate to Oxford's Department of Statistics in 2027 from New York University and the Flatiron Institute) has been awarded a five-year Turing AI Global Fellowship. He will develop a predictive mathematical foundation for scalable deep learning.

Modern AI systems, such as the large language models behind today’s chatbots, are trained at scales where experimentation is costly and scientific feedback is slow. Yet much of their development still relies on trial and error, with little advance understanding of how a new algorithm or improvement will benefit the model. Professor Wu’s Fellowship aims to change this by developing a mathematical theory of deep learning with predictive power: one that can forecast how an AI system will behave before expensive full-scale training is undertaken. 

A key aim is to develop small, inexpensive models whose behaviour can forecast that of much larger systems. This is based on the idea that the learning behaviour of AI systems is often governed by a small number of measurable quantities that mathematics can describe precisely. 

The work follows three directions: predicting how the best training settings change with model size, so they can be tuned cheaply at small scale then scaled; understanding when a new generation of training algorithms genuinely speeds up learning; and building simplified “sandbox” models that reveal how the choice of training data determines whether a model learns a reliable reasoning strategy or a brittle shortcut. Beyond new theories, the Fellowship will deliver open-source tools and benchmarks, help the UK maximize its investment in AI computing, and train researchers fluent in both mathematics and modern AI.

Portrait of Denny Wu, a man of Asian descent. In the background is Mount Fuji.
“The field of deep learning moves so quickly that, as a theorist, I sometimes feel like I’m writing the manual for a machine that has already been upgraded twice. This Fellowship gives me the resources to explore foundational questions in depth, turn the answers into tangible impact on how AI is built, and, with luck, finish the manual before the next upgrade.”
— Professor Denny Wu

More information about the new Turing AI Fellows can be found on the UKRI website.

For more information about this story or republishing this content, please contact [email protected]