Classify sensitive columns across the data catalog
Scan tables to tag columns likely holding emails, phone numbers or IDs so governance teams can review and set access rules.
The use case
Governance teams rarely know where personal data lives across thousands of tables. Databricks data classification scans Unity Catalog tables and tags columns likely to hold sensitive data such as emails, names or national IDs. Governance teams review the suggested tags and use them to drive masking and access policies. New tables get scanned as they appear. Someone still confirms tags, since columns named 'notes' may hide personal data and others may be flagged wrongly, and policy changes go through the normal approval process.
Example workflow
- Enable classification on target catalogs
- Review suggested tags by table
- Confirm or correct each tag
- Apply masking or access policies to confirmed tags
- Review new detections on a schedule
Tools
- Databricks Unity Catalog
Risks and guardrails
- Missed detections leave sensitive data exposed
- Over-tagging can block legitimate access
- Policy changes need governance approval
Sources
Written by Agents Lookup. Credits and official pages for the tools this use case names:
- Databricks docs: data classification: docs.databricks.com/aws/en/data-governance/unity-catalog/data-classification