Google just released findings from a first-of-its-kind collaboration with Stanford researchers that tackles one of enterprise tech's biggest headaches: why employees resist AI tools even when they're available. Martin Gonzalez, Org Design and Development Lead at Google DeepMind, shared five data-backed strategies that could reshape how companies roll out AI across their organizations. The timing couldn't be more critical as enterprises pour billions into AI infrastructure while struggling with single-digit adoption rates among frontline workers.
Google is pulling back the curtain on what it learned the hard way about getting employees to actually use AI tools. The company teamed up with Stanford researchers to study adoption patterns across its workforce, and the results reveal why so many enterprise AI deployments stumble right out of the gate.
Martin Gonzalez, who leads organizational design and development at Google DeepMind, published the findings on Google's official blog today. The research digs into a problem that's costing companies billions - they're building AI infrastructure faster than their people can adopt it.
The collaboration marks a rare public acknowledgment from Google that deploying AI internally isn't as simple as flipping a switch. Even at a company where engineers literally build these models, getting widespread adoption requires intentional strategy. That reality check matters for every enterprise CIO currently wrestling with utilization metrics that refuse to budge.
Google's research tackles the human side of the equation that most vendors conveniently skip over in their pitch decks. While companies like Microsoft and OpenAI race to embed AI copilots everywhere, the bottleneck isn't technology anymore - it's organizational change management.
The five strategies emerged from studying both successful and failed AI adoption attempts inside Google's walls. The research doesn't just rely on surveys - it tracked actual usage patterns, productivity metrics, and employee sentiment over time. That longitudinal approach gives the findings more weight than typical tech industry research that often amounts to glorified marketing.
