Govern the Foundry Model Supply Chain
Ensure only approved, eligible models reach Foundry projects and that every deployment meets a minimum guardrail bar — with consumption ceilings that keep usage within governed limits.
Core capabilities
-
Use built-in Azure Policy to restrict which models and deployment settings are permitted, so projects can only deploy from an approved, eligible catalog
-
Apply a baseline set of content and safety guardrails to every deployment so no model reaches production below the minimum protection bar
Supporting capabilities 1
-
Enforce token and quota ceilings through the AI gateway as consumption governance, capping how much any deployment can spend