Discoverability by Design: Data, Information and AI Jenny Brooker, Chief Data Architect at GDS Jody Parker, Head of Government Knowledge and Information Management at GDS London 2026 #Designing4Meaning #wiad26
A presentation at World IA Day London 2026 in March 2026 in London, UK by London World IA Day
Discoverability by Design: Data, Information and AI Jenny Brooker, Chief Data Architect at GDS Jody Parker, Head of Government Knowledge and Information Management at GDS London 2026 #Designing4Meaning #wiad26
How is AI reshaping user behaviour? The way people find, consume and act on information has changed forever. Large language models aren’t just new tools - they represent a fundamental shift in how people seek, consume and interact with information. Users no longer just search and browse; they increasingly converse, delegate and expect answers. This behavioural transformation has profound implications for the data systems and infrastructure that power modern organisations.
What does AI mean for Enterprise Data? Whether structured or unstructured, AI changes what your data needs to be prepared, contextualised, and reachable. Unstructured Data Structured Data The dark matter of the enterprise - Built for machines, but AI demands it abundant and largely invisible. LLMs speaks to humans too - can unlock it at scale, but only with self-describing schemas, clean data, proper governance. and lineage as infrastructure. • • Legacy access & poor permissions • Dispersed sensitivities • Missing/corrupted metadata • Duplicate/outdated content ROT Self-describing schemas & semantic layers • Clean data - dirty inputs = wrong AI answers • Lineage graphs as core infrastructure • Documentation as a first-class concern
Discoverability by Design: Metadata is sexy again Metadata is now the connective tissue between content and AI’s ability to reason over it. Tagging for Hierarchy Reliability of AI Outputs Discoverability as Architecture Structured tags help AI retrieve the right Well structured content reduces hallucination Every artefact - document, dataset, API - content - not just any content. and improves precision born with rich metadata from day one.
The one thing every IA professional must do If there is one single, urgent action for information architects, data architects, data engineers and technical product leaders - it is this: make information and data discoverable by design, not by chance.
