Are there any metadata stores that keep audit logs for model changes across teams?

Yes — a few metadata/catalog/governance platforms support audit logging for model or metadata changes across teams, though the exact depth varies.

Common options:

• DataHub
  Has rich metadata history and auditability. You can track who changed what and when for entities like datasets, schemas, and ML models. Good for cross-team governance.

• OpenMetadata
  Includes audit logs / activity feeds for metadata changes, ownership updates, tags, glossary terms, etc. Works well for team-level change tracking.

• Amundsen
  More lightweight; historically stronger on discovery than full audit trails. Usually not as strong for change auditing as DataHub/OpenMetadata.

• Collibra / Alation / Atlan
  Enterprise governance tools typically offer robust audit trails, approvals, lineage, and role-based change history across teams.

• MLflow Model Registry
  Not a full metadata store, but it does keep versioned model lifecycle changes and stage transitions, which can serve as an audit trail for model versions.

• AWS Glue Data Catalog / Lake Formation, Apache Atlas
  Can provide some metadata change tracking and governance features, though the usability and completeness of audit history depend on setup and integrations.

If you mean ML model changes specifically across teams, the best fit is usually one of:

1. DataHub
2. OpenMetadata
3. MLflow Model Registry plus an external audit log
4. An enterprise governance platform like Collibra or Atlan

If you want, I can also give you a shortlist based on:

• open source only
• best for ML models
• best enterprise governance
• best for audit/compliance