The Model Context Protocol (MCP) is one of the building blocks of AI interoperability, providing AI models with a secure way to access external data sources and services. It’s the plumbing that allows a chatbot to access your calendar, your database, or your internal tools, rather than engineers creating custom pipes for each connection. Next week, that protocol will receive a major update, and while end users may not notice it, it could make a big difference in how the ecosystem develops.

The official spec for the new version has been public since May, but on Monday morning we received an unusually clear explanation of the changes from the folks at Arcade. Basically, MCP is changing the way it handles session IDs (the little tokens that servers use to remember “ah, this is the same conversation as five seconds ago”), so that servers can more easily operate on a larger scale.

As Arcade’s Nate Barbettini says:

[Under the current system] The first time an MCP client like Claude connects to a server, it sends a “hello”: I’m Claude, here’s my version, here are my capabilities. The server responds with its own capabilities and returns a session ID… From then on, the client sends that session ID in every request so the server knows it’s the same conversation. Sometimes the DNI expires, so the client has to realize it, request a new one and continue….

Imagine a real deployment. You’re running a server for millions of users, behind a load balancer whose entire job is to route each request to whatever server in the farm is free, sometimes in a different region. Now each of those machines has to know a session ID that was delivered by another machine. It’s not impossible, but it’s a serious hassle and fights the load balancer instead of working with it.

In other words, the current setup assumes one server remembers you, but real businesses distribute traffic across dozens of servers that don’t talk to each other by default, so today’s MCP servers have to do extra work just to keep track of who’s who. This has been a major headache for anyone running an MCP server at scale, and part of the reason we haven’t seen more companies offer their own MCP integrations at scale despite all the hype around agent AI this year.

Under the new system, the protocol will take a more flexible, “stateless” approach to session IDs on the server side, similar to how most regular websites already work, which should make the entire system much easier to maintain and, in theory, cheaper to run at scale.

This is all pretty technical, but it’s an important reminder that not all parts of AI development are moving at breakneck speeds. While model training is advancing, much of the technical infrastructure those models need is still subject to the slow progress of consensus among standards bodies. It’s really happening; It’s just a little slower!