Solving the Gen AI Productivity Gap Requires a Real Culture of Innovation
In 2026, AI remains one of the most heavily funded business initiatives. Rapid public adoption, persistent executive FOMO, and—most importantly—the promise of radical efficiency gains continue to fuel investment. Yet as more reports emerge highlighting the limited ROI of generative AI (GenAI) programs, a harder question is surfacing: is GenAI actually delivering the transformation we expected?
As one senior leader at a Fortune 500 company recently told me: “Compared to every major technology shift—PCs, productivity software, cloud, mobile—GenAI may be the first in decades that’s delivering surprisingly little measurable productivity.”
The Problem Isn’t the Technology — It’s How We’re Deploying It
There’s no debate that GenAI can save time on specific tasks. We’ve all experienced how easily it drafts presentations, produces meeting summaries immediately after video calls, or enables developers to write code in days rather than months.
But time saved does not automatically translate into productivity gained. The question is: why?
The answer often begins with pilots. For years, pilots have been the fastest, safest way to test and deploy new technologies. With GenAI, however, pilots are becoming part of the problem.
Imagine a pilot that saves employees a few hours per week. First, it’s nearly impossible to accurately measure how much time is truly freed at the individual level. Second, those “saved hours” stay with the employee, and we often have limited visibility into how that time is reinvested. Even when the time is productively used, the impact doesn’t scale if the individual sits inside an isolated pilot.
By design, pilots are small and incremental. GenAI, however, demands something different. If we want it—and future AI applications—to materially transform productivity, we must move from experimentation to systemization, and from individual use cases to organizational redesign.
That means stepping back and rethinking entire business processes and workflows end to end.
We’ve been here before. When word processing tools first appeared, they primarily replaced typewriters for administrative staff. Productivity gains were real—but limited. The true transformation came later, when organizations realized the computer wasn’t just a faster typing machine. It fundamentally changed how information was created, shared, stored, and scaled across the entire enterprise.
GenAI sits at a similar inflection point. The real opportunity is not at the task level, but at the process level.
Tech Transformation Always Comes at a Cost
Any honest discussion about GenAI productivity must acknowledge trade-offs.
Much of the time “saved” through AI is now being reinvested in supervising the technology itself. Many knowledge workers currently treat GenAI the way they’d treat a junior colleague: fast, eager, occasionally brilliant—but never fully autonomous. It misunderstands context, produces errors, and requires continual oversight.
Endless prompt iteration, validation cycles, cross-checking, and rework can quickly absorb the very gains AI promised to deliver. The work hasn’t disappeared—it has simply changed shape.
There are also downstream effects on how we develop talent. For decades, early-career professionals built foundational skills—judgment, cultural understanding, institutional knowledge, and relationships—through routine work. As many of these tasks are now delegated to GenAI, organizations must rethink how junior talent is onboarded, trained, and developed. Junior onboarding and early-career learning are two business processes in urgent need of redesign—and that redesign itself claims time that comes directly out of any productivity dividend.
Finally, costs still matter. Licensing, infrastructure, and—critically—energy consumption must be factored in. One client recently compared a traditional offshore call center to a fully AI-powered alternative and found that the total cost of building and operating the AI-enabled model was significantly higher. In some scenarios, GenAI is not just a suboptimal answer—it’s the wrong answer.
Maximizing Technology ROI Requires a True Culture of Innovation
“To someone with a hammer, every problem looks like a nail.” At times, GenAI has become that hammer—applied enthusiastically, but not always wisely.
A true culture of innovation is what allows leaders to identify and prioritize where technology genuinely creates value—and where it doesn’t.
As Growth and Innovation leader at ManpowerGroup, my teams focus on deploying AI and other emerging technologies in service of our most urgent, measurable business challenges. These challenges are almost always embedded within specific business processes. That gives us a baseline for productivity and a clear hypothesis for improvement or disruption.
Once the focus is clear, we evaluate how best to drive impact and speed to market.
Sometimes that means building from the inside. Internal innovation allows us to leverage deep institutional knowledge, customer access, and existing infrastructure. But it can be slow. Legacy systems, competing priorities, and risk aversion can stretch even the best ideas into multiyear journeys.
Sometimes the answer lies outside. Startups bring fresh thinking, new business models, and energy unburdened by legacy constraints. Yet outsourcing innovation introduces its own friction: integration challenges, scalability issues, and the constant translation between startup logic and corporate reality.
The most effective approach often lies in between.
A hybrid model—balancing the realism of the core with the imagination of the edge—can accelerate meaningful progress. External partners working alongside internal teams help surface blind spots and prevent organizations from getting stuck in either technical or cultural legacy patterns.
In every case, the conclusion is the same: this is never just about GenAI or technology alone. Real productivity transformation happens at the intersection of People, Process, and Tech.
Asking the Right Questions
We start by challenging ourselves with a few core questions:
- How can we redesign processes and workflows to drive system-wide productivity while staying true to our North Star: Human First, Digital Always?
- How can we capture where time is saved, by whom, and how it is reinvested back into the organization?
- Which domain-specific workflows have the potential to scale from isolated use cases to enterprise-wide impact?
- How do we rigorously determine whether redesigned processes outperform the status quo in measurable ways?
The innovation imperative has never been stronger as leaders collectively search for real answers to the GenAI productivity question.
As leaders gather in Davos, this is exactly the conversation we need to be having—moving beyond pilots, hype, and isolated wins to reimagine how people, processes, and technology come together to create real productivity. I look forward to exchanging perspectives, learning from peers, and shaping braver, faster paths forward—together.