MCP in Production: Governing Agentic API Consumption | DeveloperWeek

As AI agents begin interacting with APIs, traditional API governance models need to evolve. In this DeveloperWeek session, Derric Gilling (WSO2) explains how organizations can manage and secure agent-driven API consumption using the Model Context Protocol (MCP).

Unlike human applications, AI agents can generate large volumes of API calls from a single prompt. Without proper controls, this can lead to unexpected costs, security risks, and limited visibility into how APIs are being used.

This talk explores how MCP provides a standardized way for AI models to discover and use APIs through structured tools. It also covers practical governance approaches for production environments, including agent authentication, fine-grained access control, cost monitoring, observability, and AI-layer security.

You’ll also learn best practices for designing APIs and MCP tools for AI agents, and how specifications like Arazzo help define machine-readable workflows.

Watch the full session to understand how API platforms can safely support AI-driven systems in production.

#wso2 #mcp