The race to deploy agentic AI is intensifying, with organizations feeling immense pressure to integrate advanced protocols like Model Context Protocol (MCP) and Agent2Agent (A2A). At Kong’s API Summit in New York, discussions highlighted the confusion among industry professionals regarding these emerging standards. Both MCP and A2A present unique complexities—MCP is frequently changing, making it challenging for developers, while A2A lacks essential telemetry and audit standards. OpenAPI remains a favored option due to its developer-friendly ecosystem, but it fails to meet the nuanced needs of AI agents with its rigid schemas. Peranich from Kong emphasized the difficulties of building secure systems under these early-stage protocols, compounded by issues in authentication and lifecycle management. The future remains uncertain—will AI protocols converge into a single standard, or will a polyglot ecosystem dominate? Maintaining oversight over various agents while enabling autonomous interaction poses an ongoing challenge for developers.
Source link
