©
Lennard Lee, NDM, Oxford University
Led from the Centre for Immuno-Oncology within Oxford’s Nuffield Department of Medicine, the AI Cancer Scientist project addresses a long-standing challenge in cancer research – that translating complex immunology into viable vaccine candidates is slow, fragmented, and difficult to scale. Developing effective cancer vaccines can take 10-15 years, in part because hypothesis generation, experimental testing, and data analysis are often separated across different teams, tools, and timeframes.
The AI Cancer Scientist aims to test whether running these steps together as a single, continuous process can accelerate progress. For the first time in cancer research, the project will seek to bring together AI models of tumour-immune recognition, automated laboratory experimentation, and sovereign UK AI supercomputing into one integrated discovery platform.
At the centre of the planned system are automated research pods. These are designed to support AI systems that generate vaccine hypotheses, design and run immune-function experiments using laboratory automation, analyse results, and iteratively refine cancer vaccine targets and formulations at scale. By integrating modelling, experimentation, and computation into a closed loop, the team will assess whether AI Scientist approaches can deliver a step change in the speed and efficiency of translating cancer immunology into patient-ready vaccine candidates.
Dr Lennard Lee, Associate Professor in Oxford’s Centre for Immuno-Oncology, practising oncologist, and project lead, said: “As a doctor, I see every day how urgently patients need better options. As a scientist, I see how slow and fragmented discovery can be. This project brings those together by creating new ways to explore cancer biology at a pace and scale that have not been possible before. We have brought together our best scientists and leading minds, and by combining AI, automation, and deep biological expertise, we aim to focus discovery where it matters most for patients.”
NDM colleagues at the launch of the AI Cancer Scientist project.
Cancer remains one of the leading causes of death worldwide, with millions of people diagnosed each year and many cancers still lacking effective preventative or long-lasting treatment options. While immunotherapies have transformed care for some patients, extending these benefits more broadly will require new approaches that can systematically uncover which immune responses are most effective against different cancers.
Commenting on why this project was selected, Ant Rowstron, ARIA's Chief Technology Officer said: “We chose to support the AI Cancer Scientist because it tackles a genuinely hard scientific world-wide problem where speed and scale are part of the solution. Cancer vaccine discovery provides a demanding real-world test of whether integrated AI systems can reason, plan, and run experiments in ways that meaningfully change how science is done.”
The project forms part of ARIA’s AI Scientist programme, a national initiative exploring whether autonomous systems can carry out the full scientific cycle under real-world constraints. Insights from the work are expected to contribute to the UK’s wider ambitions in AI-enabled scientific discovery.