Skip to content

Unlocking AI Potential with New Standardization Protocols: Empowering CIOs for Efficient AI Deployment

  • 2 min read

In the rapidly evolving landscape of artificial intelligence (AI), new standardization protocols are emerging to streamline model connections and agent management, aiding Chief Information Officers (CIOs) in devising more efficient AI deployment strategies while avoiding vendor lock-in. Notably, three new protocols—Model Context Protocol (MCP), Agent Communication Protocol (ACP), and Agent2Agent—are providing clear directions for IT leaders to move beyond failed proof-of-concept projects from the past two years and embrace measurable AI advancements.

The Model Context Protocol (MCP), released by AI company Anthropic in November last year, aims to provide a standardized approach for connecting AI models across different data sources and tools. Its primary advantage lies in allowing users to flexibly choose the best-performing language model (LLM) and its supplier, steering clear of vendor lock-in dilemmas. Dubbed the "plumbing system," MCP effectively connects various AI components. Moreover, Microsoft and other AI suppliers have begun to support MCP, enhancing its compatibility.

Following suit, IBM launched the Agent Communication Protocol (ACP) this year, enabling AI agents from different vendors to interconnect. Utilizing standard HTTP communication patterns, ACP simplifies the integration process, offering higher interoperability and reusability for enterprises. In the same month, Google introduced the Agent2Agent protocol, further promoting collaboration among diverse AI agents. With support from over 50 technical partners, Agent2Agent enables businesses to chain a series of AI agents, more easily accessing the specialized functionalities they require.

With the advent of these protocols, experts foresee the birth of AI agent marketplaces, allowing users to choose from a plethora of pre-built agents or models from various suppliers without the hassle of training their own models. These new protocols signify a new path for enterprises towards scalable AI adoption, making standardization a key driver for rapid development. As Christian Posta, Global Chief Technology Officer at Solo.io, aptly puts it, "Speed without standardization leads to chaos, while standardization provides a safeguard for purposeful expansion."

Leave a Reply

Your email address will not be published. Required fields are marked *