Five years ago, digital transformation inside large enterprises followed a familiar pattern. Technology initiatives were introduced centrally, often by IT or digital teams, and pushed into operating environments measured on uptime, cost and output. Digital was viewed as organizational infrastructure — sometimes a cost center. Adoption had to be justified, negotiated and, at times, enforced.
Earlier in my career, working on enterprise digital initiatives in manufacturing environments in Asia Pacific, that tension was explicit. When a new system threatened even minutes of downtime or added administrative load without shifting incentives, resistance wasn’t just common — it was rational.
One plant manager, half-joking as he pushed back, distilled the calculus succinctly: “For the cost of this digital implementation, I could hire 10 more operators and guarantee my output. Why should I bet on the tool?”
But was it really resistance?
I’ve seen this pattern repeat across plants and projects. In one Asia-Pacific manufacturing rollout I worked on, adoption stalled not because the technology failed, but because the new tools shifted risk and accountability faster than incentives could adjust. What looked like resistance from the outside was, in practice, rational behavior inside the system.
I came to understand later that what leaders often label as resistance is usually something simpler: a lack of clarity. The system was behaving exactly as designed.
- Downtime was punished
- Risk was localized
- Learning was abstract
In that context, opting out wasn’t defiance. It was logic — a familiar failure mode in organizational behavior research on why companies often know what to do but don’t do it.
At the time, that reasoning was sound. Digital was evaluated through a headcount-replacement lens. If technology couldn’t immediately outperform human labor, it was treated as a burden rather than a lever.
Around the same time, the language inside IT and digital leadership was beginning to change. In town halls and internal forums, the emphasis shifted from technology as a support function to IT as a driver of business outcomes rather than an enabler. I remember hearing this shift play out repeatedly in leadership conversations and IT town halls, even as day-to-day behavior on the floor remained unchanged. The rhetoric was shifting — even if behavior on the floor had not yet caught up.
That era is over. What replaced it didn’t arrive through a program or mandate. It arrived quietly, unevenly and much faster than most organizations realized.
When technology moves from mandate to habit
Across consulting engagements, academic settings and recent internship environments, a fundamental behavioral shift is becoming visible. AI is no longer something employees wait to be given by IT. It is something they actively pull into their own work.
Individuals are embedding enterprise AI tools directly into daily workflows — developing prompts, sharing templates, building chatbots for specific tasks and, in some cases, creating narrowly scoped custom models. This activity is rarely mandated. It spreads informally, through peer exchange rather than program management. Much of it resembles what IT leaders once labeled shadow IT, a term long used to describe technology adoption that occurs outside formal governance frameworks.
The most striking change is not the sophistication of the tools, but the shift in agency. For years, leaders tried to drive change through abstract direction: “be more digital,” “use AI responsibly,” “innovate faster.” Those messages failed not because people disagreed, but because they offered no concrete guidance on what to do differently on Monday morning.
AI breaks that pattern by making experimentation cheap and legible. When individuals can test ideas, compare outcomes and reuse what works with minimal friction, adoption becomes a personal decision rather than an institutional one. This is why employee use is now growing ahead of organizational implementation, even as formal rollout and governance lag. A prompt that saves two hours doesn’t need translation, sponsorship or persuasion. Its value is self-evident.
In past transformations, leaders tried to persuade people to change. AI flips that dynamic. It lets individuals run their own low-cost experiments, generate their own evidence and adopt based on results rather than rhetoric. For the first time, the burden of proof has flipped. Instead of leaders having to convince the organization, individuals generate their own evidence through use — and evidence travels faster than arguments ever did.
The 10-man parity rule
This shift can be understood through a simple heuristic: the 10-man parity rule.
In earlier waves of digital transformation, technology adoption was implicitly benchmarked against human labor. A million-dollar system had to justify itself by replacing people. In capital-intensive environments, where reliability matters more than experimentation, that hurdle was high.
Today, the calculus has flipped. The rule is no longer about replacing 10 people; it is about enabling one person to operate with the leverage of 10. The concept is not headcount math — it reflects how organizations price risk and agency at different stages of technological maturity. Once one person can produce the output of 10, budgeting, risk review and performance evaluation all become lagging indicators rather than controls. At early stages, leverage accrues to people who question assumptions, experiment quickly and connect ideas across domains — behaviors long associated with innovative performance.
When an employee realizes that a prompt, agent or chatbot can compress weeks of analysis into hours — whether for compliance, deal execution or operational planning — adoption becomes a personal decision. There is no need to wait for a rollout plan. This is why AI adoption is now consistently outpacing formal governance structures, a gap highlighted repeatedly in recent industry research.
What distinguishes this pattern is that it rarely announces itself as “innovation.” It starts small, in corners of the organization that don’t trigger investment thresholds or governance alarms. By the time leadership recognizes it as strategic, the behavior has already spread.
Viewed from a strategic lens, this is not just process improvement. When individuals gain disproportionate leverage through AI, the implicit rules of the organization begin to shift — who creates value, what work matters and how decisions are shaped. Those changes surface first in behavior, not structure. By the time titles or role definitions catch up, the real reconfiguration has already happened.
Three second-order effects leaders are underestimating
What makes the shift from mandate to pull strategically different is not adoption, but acceleration. When AI is embedded individually rather than deployed institutionally, learning cycles compress unevenly across the same organization.
Silent acceleration
Capability now compounds faster than formal systems can observe. Some individuals rescope their roles, internalize new workflows and move through learning curves at a pace that performance frameworks neither track nor reward. Others fall behind — not through resistance, but through inertia. I’ve watched this divergence emerge quietly inside teams that, on paper, look identical. The result is widening performance divergence without formal signals. Formal management systems are designed for stability — to surface failure and cost — not to track learning velocity.
What makes this divergence hard to spot is that it doesn’t show up in headcount, budgets or org charts. It shows up in informal signals — who people go to for answers, whose work gets reused and whose judgment starts shaping decisions. Speed and clarity get rewarded long before formal systems catch up.
Over time, this capability gap doesn’t stay invisible. It begins to reshape how decisions are made. Left unmanaged, this doesn’t just widen performance variance — it distorts how work gets staffed, who gets trusted and which insights make it into decisions.
Shadow promotions
As AI leverage accumulates informally, influence migrates before titles do. Individuals who can frame problems, synthesize insight or automate complexity become the default path to clarity — even without formal authority.
These so-called shadow promotions harden into real power over time — often before organizations recognize or legitimize them. In several cases I’ve observed, authority shifted months before any formal change followed.
This kind of tension is predictable. New ways of working almost never sit comfortably alongside old ones. When AI amplifies individual leverage, it collides with existing incentives, reporting lines and notions of merit. Influence shifts quietly first. Formal recognition, if it comes at all, follows later.
Rate-of-change mismatch
The hardest coordination problem is no longer functional alignment but reconciling uneven speeds of learning. Teams operating at different rates of AI fluency struggle to collaborate, not because of incentives or intent, but because they are effectively working at different cognitive tempos.
Traditional management systems assume uniform progression — an assumption AI quietly breaks. This rate-of-change mismatch is now one of the most persistent AI governance challenges leaders are grappling with. The result is coordination drag: meetings slow down, standards fracture and teams start arguing about methods instead of outcomes.
From rollout to governance
This bottom-up adoption changes the texture of enterprise transformation. Innovation no longer arrives as a centralized program; it emerges in fragments. Practices evolve faster than formal structures. Knowledge spreads laterally rather than hierarchically.
I’ve watched teams adopt AI tools faster than IT roadmaps could absorb them. In multiple engagements, analysts and managers were already using copilots, scripts and automation quietly — not because IT failed, but because performance expectations moved faster than formal approval cycles. By the time governance discussions surfaced, behavior had already changed.
In this environment, the role of the CIO and digital leadership has quietly shifted. The challenge is no longer persuading reluctant operators to adopt tools. It is governing the enthusiasm of a workforce that is already moving faster than the IT roadmap. Governance replaces rollout as the primary strategic function.
In practice, that tuning shows up in small but consequential choices:
- Which uses are quietly tolerated versus explicitly shut down
- Which AI outputs are allowed to circulate without review
- Which teams get air cover to experiment and which are told to wait
None of these decisions look strategic in isolation. Together, they determine whose capabilities compound — and whose quietly stall — long before budgeting cycles, risk reviews or performance systems notice.
An uncomfortable truth
Leaders must confront an uncomfortable truth: in the AI era, some internal inequality is inevitable. The strategic choice is not whether it emerges, but whether it is shaped deliberately or allowed to form by accident — rewarding those who bypass governance while quietly losing those who outgrow it.
The risk is that these changes look minor at first — informal, fragmented, easy to dismiss. Early signals rarely fit established metrics or planning cycles, so they get rationalized away. AI’s bottom-up adoption follows this pattern closely. By the time the impact becomes undeniable, uneven capability has already solidified into structure.
This is the inflection point for CIOs. The job is no longer to sell ideas, build consensus or orchestrate adoption. That work is already happening without permission. The new responsibility is governance — not of tools, but of experiments already in motion. Who is running them, where they are compounding value and where unmanaged divergence becomes risk.
The question is no longer whether experimentation should happen. It is whether leadership is aware of it early enough to shape the outcome.
A new phase of digital reality
The defining feature of the AI era is not the power of the technology itself, but the shift in who brings it into the organization. Digital transformation has moved from being pushed by institutions to being pulled by individuals.
The first time I realized how far ahead reality had moved was during a formal performance review cycle, when outcomes were credited to workflows enabled by tools that technically “did not exist” in our official stack. The org chart still said one thing. The operating model said another.
For leaders, the strategic question is no longer whether people will use the tools. That question has already been answered. The real challenge is whether organizations can recognize — and respond to — the silent acceleration already underway, before informal advantage hardens into structural inequality.
This article is published as part of the Foundry Expert Contributor Network.
Want to join?
Read More from This Article: The 10-man parity rule: When AI adoption accelerates faster than organizations can see
Source: News