The ODRL journey – finally making Access machine-actionable

Abstract

In 2026, applying for access to data is still more complex and challenging for researchers than it should be. To some extent, this is because access & usage policies are still largely mediated by humans, compounded by a lack of alignment between organisations as to how access processes and workflows are implemented, with little supporting machine-actionable information to improve the situation through automation. This not only degrades the researcher experience, it inhibits progress on federating infrastructures for data that is not completely open. Open Digital Rights Language (ODRL) has been around for over a decade but is now gaining significant traction in a number of arenas as a mechanism to address the problem of machine-actionable access, usage, and rights management more generally. The recent and rapid advent of agentic AI is now making rights management for digital objects a priority issue that can no longer be deferred. “Access” is one of the key planks in the Cross-Domain Interoperability Framework (CDIF) and builds out implementable specifications for the “A” in FAIR. This presentation outlines the work that the UK Data Service has so far undertaken to move ODRL from a standard to an implementable set of metadata, workflows, and tools. We also articulate in detail how this is not merely a simple exercise in translating prose licences to RDF but requires a deeper understanding of the context in which “Access” operates. We also describe how standards such as Data Use Ontology (DUO) and Data Privacy Vocabulary (DPV) complement ODRL within that context.