What's new in Collibra July 2026

AI is moving fast, but trust hasn't kept up. In this release overview, Michael Rahm walks through everything new in Collibra's July 2026 launch, designed to make Collibra the foundation for scaling AI you can actually trust. What's new: Governed Context Compiler – Deliver structured, trusted context (metrics, definitions, policies) from the Collibra Knowledge Graph to any AI agent or platform — including via MCP server at runtime. One governed source of truth, no custom engineering. Collibra Data Access (GA) – Centralize entitlements, permissions, and column masking across Snowflake, Databricks, and BigQuery from a single interface. Faster access, consistent policies, compliance intact. Agent Dashboard & AI Trust Score – A dynamic 0–100 trust score for every registered agent, with trend lines, evaluation coverage gaps, and native Databricks Agent Bricks integration — all in the AI Command Center. Automated AI Traceability for Snowflake Cortex AI – Connect agents, prompts, semantic views, and datasets directly to Collibra's AI Command Center. Data Product Lifecycle Manager – Standardize how every team builds data products with gated stages, validation, approval, and publishing — so "trusted" means the same thing every time. Data Product Integration – Native ingestion of data products from Snowflake, Databricks, Google Cloud, and SAP Datasphere, all governed under one standard. SAP Signavio Integration – Bridge process documentation and governed data. Trace accountability from a business operation all the way to the underlying data asset — essential for DORA, BCBS 239, GDPR, and auditable AI. 🔗 Learn more: collibra.com/innovations | docs.collibra.com