Ryan Lufkin and Melissa Loble step back this episode and hand the conversation to two technologists who interview each other. Mike Mast, Principal Group Program Manager for Microsoft Education and a 1EdTech board member, and Zach Pendleton, Chief Architect at Instructure, trade questions about what interoperability means once AI agents start acting inside the tools educators and students already use.

The starting point: when the client is an agent rather than a person, agents can learn new vocabularies and reason over different systems on their own, so the hard part shifts from rigid data schemas to giving agents a real semantic understanding of a course. Mast points to Canvas by Instructure and its Smart Search beta as one of the few early examples of that idea in practice. Pendleton makes the case that none of this replaces the open standards ed tech already runs on. He frames MCP, LTI, OneRoster, and Caliper as an "and," not an "or."
From there the two get into the part that keeps Pendleton up at night: as actions move further from the human who set the goal, how do we keep people in authority? They cover delegated authorization and consent, audit trails, the gap between what AI can do and what we want it to do, and who carries the cost when an automated tool gets it wrong.

In this episode:

Agentic AI changes what interoperability optimizes for, moving the priority toward semantic understanding of course content rather than identical schemas across systems.
New protocols build on existing standards instead of replacing them, so prior investments in open APIs and LTI still matter.
Keeping a human in control sometimes costs a little speed, and both guests argue that maximum speed should not be the goal of an educational AI system.
Transparency is non-negotiable: if AI touches a grade, students should know, and they should be able to question the result.