AIRR-BRIDGE: Bridging Research Infrastructure for Data Governance Extension on AIRR

Training advanced AI models increasingly requires access to powerful graphics processing units and large-scale computing systems. However, these resources have historically been difficult to integrate with the TRE infrastructure designed to protect sensitive data.

This project will federate TREs at The Alan Turing Institute and King’s College London with the UK’s AI Research Resource (AIRR) supercomputers, including the Dawn system in Cambridge and the Isambard-AI system in Bristol.

Through exemplar studies in cardiac imaging, breast cancer pathology, and brain imaging, the project will test technical and governance approaches to enable secure analysis of sensitive health data on the national supercomputing infrastructure.

Public involvement activities will focus on building trust and transparency in the use of sensitive health data within national AI systems, helping shape guidance for future large-scale research infrastructure.

By the end of the project, AIRR-BRIDGE will:

• Demonstrate secure federation between TREs and AIRR supercomputers
• Enable analysis of UK Biobank and KCL datasets on the national AI infrastructure
• Evaluate satellite TRE extensions to supercomputing environments
• Assess governance alignment with the SATRE TRE standard
• Develop a reusable blueprint for scaling secure access to AI supercomputing

Project information

Lead organisation: The Alan Turing Institute
Principal investigator: Martin O’Reilly
Project duration: 12 months
Project partners: King’s College Hospital NHS Foundation Trust, University College London, University of Bristol, University of Cambridge, UK Biobank
Funding provided: £487,300
Primary contact email: moreilly@turing.ac.uk