Skip to main content

Search the Commons

Type at least 2 letters. Use the arrow keys to move through results, Enter to open, Escape to close.

Ask a question instead: answers drawn from across the Commons, with links to the source. Or browse the site map.

Manage and improve · AI for data quality

Getting real value from AI to improve data quality, safely

Guidance, use cases, worked examples and case studies for using AI on data quality, whether it is built into a supplier's product or built as your own tool. Also: the route by which an AI capability is tested and assured before it joins the shared toolkit for everyone to reuse.

In short: a plain English summary
  • This page is about using artificial intelligence (computer programs that learn) on data.
  • It shows where it helps and where it is risky.
  • People must stay in charge.

The core idea: AI drafts, finds and explains; a deterministic, reviewable rule decides. Keep people in charge of anything that changes data or affects someone.

Principles

Use cases

Use caseDimensionsMaturityHowRisk

AI inside suppliers' tools: questions to ask before you buy or switch on

Many data quality and catalogue products now include AI features. Ask these questions, in writing, and keep the answers with your DPIA.

    Then compare options on Tooling, and use the tooling survey to learn from organisations already using them.

    Worked examples

    Illustrative These examples show the method with realistic but invented details. They are not reports of real projects.

    Case studies

    Have a case study? Contribute it, or use the case study drafter on Case studies.

    From experiment to shared toolkit: the route

    Checklist before reuse

      AI and automation components in the toolkit

      ComponentKindStageOwner

      Want to take a component through the route? Raise a commission or add it to the wishlist.