Linking generative AI models to the agents that are going to use them now has a standard for queries, the Model Context Protocol (MCP). Suddenly all sorts of things are vying to be your data source of choice. What could go wrong? Analysts Jean Atelsek, James Curtis and Henry Baltazar join host Eric Hanselman to provide an introduction to what MCP is, what all the fuss is about and where this is headed. Since Anthropic released the standard in November of 2024, there has been significant progress in expanding and clarifying it, but it is still very much a work in progress. Database providers and storage vendors have been the obvious participants and many others are joining in. The larger questions are around ensuring trust and transparency in this market. Basic authentication has been worked out, but more sophisticated authorization mechanisms need to be defined.

The mad dash is already leading to M&A activity, with storage vendors looking to expand their opportunities and maintain their relevance. Microsoft has released its MCP definition and Google has created an agent-to-agent (A2A) protocol for direct agent interactions. There are many pieces that are coming together to enable fully agentic operations and there is still a lot of work to be done.

 

More S&P Global Content:

• The 2025 Generative AI Outlook

For S&P Global subscribers:

• Technology Primer: Model Context Protocol explained
• Databases and analytic services get the agentic AI treatment at Google Cloud Next 2025
• IT Insider 3: A roundup for IT decision-makers

Credits:

• Host/Author: Eric Hanselman
• Guests: Jean Atelsek, James Curtis, Henry Baltazar
• Producer/Editor: Adam Kovalsky
• Published With Assistance From: Sophie Carr, Feranmi Adeoshun, Kyra Smith