Securing Golden Data for AI Governance Metrics on Databricks A developer outlined a set of Unity Catalog controls for securing the golden data behind AI governance dashboards on Databricks, including a dedicated schema with restricted MODIFY grants, lineage queries against system.access.table_lineage to catch eval-set leakage into prompts or retrieval indexes, audit-log monitoring of permission changes, and Delta time-travel versioning so every reported metric can be tied to a specific data version. The approach treats the number of principals able to edit the answer key as a governance metric itself, and flags reads from interactive notebooks or copies to tables outside the evaluation results schema as warning signs. AI governance dashboards report accuracy, groundedness and safety scores. Every one of those scores is a comparison against golden data: the eval sets, expected answers and reference data your team has agreed to treat as truth. If that data can be quietly edited, leaked into prompts or lost to retention, the dashboard still shows numbers. They just stop meaning anything. This post walks through the controls I use to secure golden data on Databricks, with SQL you can adapt. Everything here uses native Unity Catalog features. Put golden data in a dedicated schema so you can govern it as one unit. Readers get SELECT. A small owners group gets MODIFY. CREATE SCHEMA IF NOT EXISTS ai gov.golden COMMENT 'Golden eval sets and reference data. Changes need review.'; GRANT USE CATALOG ON CATALOG ai gov TO eval-readers ; GRANT USE SCHEMA, SELECT ON SCHEMA ai gov.golden TO eval-readers ; GRANT USE CATALOG ON CATALOG ai gov TO golden-data-owners ; GRANT USE SCHEMA, SELECT, MODIFY ON SCHEMA ai gov.golden TO golden-data-owners ; Then measure it. The number of principals that can change the answer key is a governance metric in its own right. SELECT grantee, privilege type FROM system.information schema.schema privileges WHERE catalog name = 'ai gov' AND schema name = 'golden' AND privilege type IN 'MODIFY', 'ALL PRIVILEGES' ; This query only covers schema-level grants. Also check table-level grants, catalog-level grants and object owners, because owners and admins can change data too. Leakage is the quiet killer. Someone pastes eval questions into a prompt, or a pipeline copies the set into a retrieval index. Scores go up, and the model has effectively seen the answers. Unity Catalog writes table lineage to system.access.table lineage , kept for 395 days by default. Each row records the entity notebook, job, pipeline, query or dashboard , who generated it, and the source and target. A row with a source and no target is a read. This query lists every read of the golden schema in the last 30 days by anyone other than your evaluation service principal: SELECT event time, created by, entity type, entity id, source table full name, target table full name FROM system.access.table lineage WHERE source table catalog = 'ai gov' AND source table schema = 'golden' AND event date = date sub current date , 30 AND coalesce created by, '' < '