From UX to AX: Change Management for the Agentic Experience

Greg Isenberg’s post on moving from traditional User Experience (UX) to Agentic Experience (AX) really got me thinking, especially from a change management perspective.

As technical teams implement MCP (Model Context Protocol) to enable AI agents to talk and collaborate seamlessly, success isn’t just about connecting systems. It’s about how our data, context, and knowledge are structured and curated so agents can truly understand, infer, and coordinate.

This means helping business users and data owners prepare information thoughtfully, not just for human consumption, but for AI agents who rely on rich, trustworthy context to act autonomously and effectively.

Our researcher mindset,  interviewing, observing, and uncovering real user and agent needs becomes essential. Designing AX means understanding how AI agents experience systems and data, where humans remain “in the loop” or “on the loop,” and how we build trust in agent decisions.

Change management here is about enabling shared agency with humans and AI agents learning to work together as partners, with empathy for both technical challenges and human concerns.

When done well, the shift to AX and MCP-powered agent collaboration will amplify human creativity and productivity… not replace it.