Emerging Protocols Could Standardize AI Integration

• Model Context Protocol and Agent2Agent are emerging protocols for integrating AI-based agents, while NLWeb simplifies adding natural language chat to Web sites.
• Organizations can leverage the protocols to build and integrate AI applications and improve Web site usability.
• These are fledgling protocols, so planning and investments in them should be done with caution.

Model Context Protocol (MCP), Agent2Agent (A2A), and Natural Language Web (NLWeb) are new cross-vendor, cross-platform protocols that could speed the integration and capabilities of AI agents and help organizations create solutions that are driven by natural language inputs. However, all three are less than one year old and are expected to evolve, assuming they get traction.

Benefits of Standardized AI Agent Protocols

Microsoft’s latest AI initiatives are focused on creating “agentic” applications, where specialized AI assistants (agents) work together to provide information and complete complex tasks requested by users via natural language. Distributing the work among agents allows each to be highly tuned to work with data and perform tasks within a specific domain, rather than having a single high-level agent handle all work on its own. The approach is similar to the way microservices are used to decompose monolithic applications into discrete services.