Data Governance

Data Governance is the framework of policies, roles, and controls that determines how an organization's data is classified, secured, accessed, retained, and proven compliant. It answers ownership questions rather than technical ones: who may see a table, how long a record is kept, where personal data may be stored, and who signs off when those rules change. The practical output is a catalog of data assets with named owners, access rules enforced by the platform, lineage showing where each field came from, and an audit trail regulators can inspect. Tools such as Collibra, Alation, Microsoft Purview, and open-source OpenMetadata sit in this layer, while cloud warehouses like Snowflake and BigQuery enforce row- and column-level policies. Regulation drives adoption: GDPR set the baseline for personal data, and the EU AI Act pushed governance into model training data by requiring documentation of what a system was trained on. AI raises the stakes because a model trained on ungoverned data can leak or memorize it, and there is no clean way to delete one record from trained weights. The pitfall is governance as paperwork — policies that live in a document nobody enforces in the pipeline.