Microsoft has released a dedicated AI Gateway tier for Azure API Management in public preview, shifting the control plane toward models, MCP servers, and tools.
InfoQ reports that the tier can front services including Foundry, Bedrock, Vertex AI, and OpenAI behind one endpoint. Instead of relying on traditional XML policy configuration alone, the new experience uses policy cards designed around AI workloads.
The change reflects how enterprise integration is moving. AI applications often call several models and tools, each with different costs, limits, and governance requirements. A gateway can centralize routing, observability, access policy, and enforcement before requests scatter across providers.
The open question is where the governance boundary should sit. A gateway can control traffic and policy at the edge, but applications still need domain-specific checks, data handling rules, and evaluation. Microsoft’s preview gives architects a more AI-native management layer, but it does not remove the need to design responsible workflows.