What differs
- Inventory: Collibra offers an AI use case registry with data-to-model lineage; OneTrust offers an AI inventory that catalogs systems, models, agents, datasets, vendors, and projects with dependencies.
- Risk: Collibra lists risk ratings and a per-system AI Trust Score; OneTrust lists templates for EU AI Act, NIST AI RMF, and ISO 42001 with automatic risk tiering and revalidation on material change.
- Named ML integrations: Collibra lists SageMaker, Azure ML and AI Foundry, Vertex AI, Databricks, SAP AI Core, and MLflow; OneTrust lists Amazon Bedrock, Microsoft Foundry, Google Vertex, and Databricks Unity Catalog.
- Published security: Collibra’s Trust Center lists SOC 1 and SOC 2, ISO 27001, ISO 42001, and FedRAMP; OneTrust’s Trust Center lists SOC 2 Type II, ISO 27001, 27701, and 27017, and does not list ISO 42001 or FedRAMP.
Who each tends to fit
Data governance teams already on Collibra, or teams that need lineage as evidence, may look first at Collibra. Privacy and third-party risk teams already on OneTrust may look first at OneTrust. Confirm module licensing for both.
Last checked: . No winner declared.