Coffee Talk logo: a moka pot and coffee cup

The Model Context Protocol, or MCP, may be the future of integrating AI into enterprise architecture — but even if it’s not, the fact that the question is even being asked is a welcome sign of maturity in this market.

If you’re responsible for IT at any level, you probably have concerns about “shadow AI”, as every one of the tools and platforms that your users rely on adds GenAI capabilities as fast as they can release them. The initial reaction was to try to ban the use of ChatGPT and its ilk, but the realisation is starting to sink in that it’s better to give users safe ways of doing what they need and want to do.

MCP, A2A, and other emerging protocol specs are probably not the final answer to how to apply some governance atop everything that’s going, but something is definitely needed, and it’s great to see the conversation happening.

This video is also available on LinkedIn, if you prefer to join the conversation over there.