ODI_report_hero
ODI_report_hero
ODI_report_hero
ODI_report_hero
ODI_report_hero

The state of AI-ready data in local government 

New insights from UK councils on standards, readiness and reform, and how public data is evolving to support search, machine learning and generative AI. 

Service

Digital Transformation

Industry

Government

The councils that modernise their data will modernise their services fastest. Across the UK, local authorities are discovering that data is no longer a by-product of bureaucracy but the raw material for automation. When structured for search, prediction and conversation, it cuts costs, sharpens policy and powers a new wave of citizen-centred services. 

Local governments’ datasets are no longer about accountability alone; they are becoming the fuel of automation. 

Where readiness meets results 

Ten case studies, from Dorset’s predictive budgeting to London’s fire-risk modelling, reveal both progress and fragmentation. Most councils now publish open data, but few prepare it for reuse. Machine-readable metadata remains scarce, and standards are patchy. Where they do align – ISO dates, UPRNs, SNOMED CT – AI tools deliver faster insights, more accurate forecasts and measurable savings. The evidence is clear: readiness is not about scale or budget, but about the discipline of standards and governance. 

These variations point to a deeper truth: AI is not a single capability but a family of tools, each with its own data demands. 

alt=""
alt=""
alt=""
alt=""
alt=""

Different AIs,
different data demands 

AI readiness is too often treated as a box to tick. In reality, what makes data “fit for AI” depends on the purpose. The datasets that drive a chatbot will fail a forecasting model; what works for search may break predictive analytics. For councils, recognising these distinctions is essential to making sound investments. 

Efficiency and democracy are not competing goals 

Efficiency and democracy reinforce one another.

The white paper argues that operational performance, state capacity and public trust, core measures of efficiency, depend on the same values that sustain democracy: transparency, accountability and pluralism.

Centralized systems may deliver quick wins but often weaken resilience and trust. Federated, modular designs take longer yet build adaptability, sustainability and sovereignty. 

From data maturity to data purpose 

ODI and Nortal propose a framework that treats readiness as a toolkit, not a ladder:

  • Search readiness: clean metadata, canonical identifiers and machine-readable schemas.
  • Machine-learning readiness: structured, balanced data at scale with clear provenance.
  • Generative-AI readiness: annotated corpora and APIs built for retrieval and conversation.

Search readiness underpins transparency and efficiency; machine-learning readiness supports foresight and planning; and generative readiness opens the door to new forms of citizen interaction. Each requires its own standards, governance and infrastructure, but together they give councils flexible pathways to modernise services and build trust.

These are not stages of maturity but lenses of purpose – transparency, foresight and trust – giving councils practical routes to modernise services without waiting for perfect systems. 

ODI_report_book_mock

Decisions make data ready, or not 

Every dataset reflects a decision — about structure, standard, governance or neglect. Those choices shape whether councils can anticipate crises or merely record them. The report offers a toolkit for turning data into capability: adopt standards, publish metadata, automate pipelines and separate operational from strategic data. 

Get in touch

Learn more about how to build a human-centered government, provide a better public service experience and efficient public administration.