GovernanceCore
Operations

Data governance

Working definition

Reviewed 30 July 2026

Data governance is the system of ownership, rules, controls, and evidence used to manage data across its lifecycle. For AI, it covers the suitability, provenance, representativeness, quality, rights, security, access, labeling, lineage, and monitoring of training, validation, testing, and operational data.

Context

Why it matters

Data choices shape model capability, bias, privacy, security, and performance. Weak lineage also makes defects difficult to trace and decisions hard to defend.

Operating note

What this looks like in practice

  1. 01Record provenance, permissions, collection context, transformations, and known limits.
  2. 02Test relevance and representativeness against the system’s intended population and use.
  3. 03Monitor operational data for drift, quality failures, and prohibited content.

Sources & further research

Primary authority anchors the definition. Research links add conceptual or operational depth. External sources may update independently; always verify legal duties against the current official text.

  1. Official source

    Regulation (EU) 2024/1689, Article 10

    Specifies data and data-governance practices for high-risk AI systems.

    European Union
    2024    Open source ↗
  2. Research paper

    Datasheets for Datasets

    Introduces structured documentation for dataset motivation, composition, collection, use, and maintenance.

    Gebru et al.
    2018    Open source ↗