Enterprise AI rollouts are less about adopting a new technology, which we’ve all done many times before, and more about humans being asked to completely change how they approach work. This case study goes inside two rollouts at JLL, a 115,000-employee global commercial real estate firm. First, the launch of “JLLGPT", JLL's internal company-wide generative AI platform, and the 20,000-person Claude deployment that followed. From his vantage point, Dale shares what the first rollout revealed about adoption, understanding, and skill gaps, and how those lessons reshaped the second, including agentic AI that now answers live RFPs in production. Attendees will leave with a practical picture of the human learning curve behind enterprise AI: what people must unlearn, what enablement teams tend to get wrong about "training users on AI”, and how to build tools that work in BAU.