Blocked Agency
Microsoft surveyed 20,000 knowledge workers across 10 countries for this year’s Work Trend Index.
The headline framing: “Workers are ready. Their organizations aren’t.”
There’s a data point inside the report that’s more specific than that. One that names the actual problem.
Ten percent of AI users are in what Microsoft calls “Blocked Agency.” These are workers with strong AI capability — they’ve built skills, redesigned workflows, done the work. Their organizations can’t absorb what they’re capable of.
Not because the workers are wrong. Because the system around them isn’t built to hold the output.
The 67-32 split
Here’s the finding that reframes the entire debate.
Microsoft analyzed what factors actually drive AI impact — not adoption rates, not usage frequency. Real, measurable outcomes: higher quality work, better collaboration, expanded capacity, career trajectory.
The result: organizational factors (culture, manager support, talent practices) account for 67% of AI’s real impact. Individual mindset and behavior: 32%.
The organization is more than twice as responsible for AI outcomes as the individual using it.
Sit with that.
Every AI tool on the market is built for the individual. Your skills are the variable. Your prompting ability is the constraint. Your adoption rate is the number that gets tracked.
Microsoft just showed that’s the wrong layer.
What “blocked” actually means
The 10% in Blocked Agency are a specific kind of failure. Not resistant workers. Not late adopters. The capable ones — the ones who did exactly what they were told to do. They learned. They built. They changed how they work.
And they’re stuck.
Their outputs can’t be routed, governed, or absorbed by the organization around them. Their work lands, and then it sits. Without the coordination infrastructure to integrate what they produce into the broader system, their capability generates friction instead of value.
This is the inverse of what most organizations are managing for.
Most leaders are worried about the bottom: workers who won’t adopt AI. The laggards. The resisters.
The structural failure is at the top. The people who moved fastest are now waiting on a system that was never built for what they created.
What Microsoft calls it vs. what it is
Microsoft names the gap the “Transformation Paradox”: workers are ready, organizations aren’t. The prescribed fix: leadership alignment, culture redesign, manager support, talent practice reform.
That’s a change management answer to a coordination infrastructure problem.
When 10% of your most capable AI users are blocked, the issue isn’t that the culture hasn’t caught up. The issue is that there’s no system above the individuals that knows what’s open, what changed, what matters, and what needs to close.
Culture can’t do that job. A manager can’t hold that state. An alignment initiative can’t track it.
Coordination infrastructure can.
The moment AI-capable workers start generating significant output — more drafts, more analyses, more decisions, more proposals — the system that absorbs that output needs to exist. Not the people who approve it. Not the culture that welcomes it. The actual system: persistent context, commitment tracking, routing logic, loop closure.
That’s what’s missing in the 10%.
The math only works if something coordinates above it
Only 19% of workers are in Microsoft’s “Frontier” zone — the overlap of individual capability and organizational readiness. The rest are either stuck in the emergent middle, unclaimed by organizations without individuals to match, or blocked.
This isn’t a slow adoption curve. It’s a structural mismatch.
The employees moved. The systems didn’t.
And the systems didn’t move not because of culture lag, but because there’s nothing to install. The coordination infrastructure for AI output — the layer that routes, tracks, closes loops, and maintains context across everything — doesn’t exist as a default in most organizations.
It has to be built. Deliberately. As infrastructure.
Not as a culture initiative. Not as a change management program. Not as a reorg.
As a coordination layer — the same way organizations build security infrastructure, compliance infrastructure, data infrastructure. Not because someone got inspired, but because the alternative is paying the coordination tax in human attention.
The 67% is telling you where to build.
The 10% is showing you the cost of not building it.
Eliran Keren — Founder of Deeplica, building the coordination layer for knowledge work.
Source: Microsoft 2026 Work Trend Index — Agents, human agency, and the opportunity for every organization
