The learning organization: How to accelerate AI adoption

The dizzying speed at which AI technology is evolving makes it nearly impossible to keep up with the many new ways that it could transform how people work. Yet for most organizations, the gap between what’s possible and what’s implemented is steadily widening. A 2024 McKinsey Global Survey found that nine in ten employees used gen AI for their work, and 21 percent of them were heavy users.¹ But while employee enthusiasm was high, the formal adoption of AI tools across most organizations lagged behind: Only 13 percent of surveyed employees considered their organization to be an early adopter.

The slow institutional embrace of new tools isn’t a new phenomenon. It happened when digital natives collaborated using cloud-based platforms and connected with customers via social media years before their employers officially approved these technologies. More recently, mobile natives started conducting business through messaging apps and mobile-first workflows while corporate IT departments were still debating smartphone security policies. Now the world is witnessing the emergence of AI natives—typically, younger employees—who are already using gen AI tools to draft emails, write code, and analyze data, while decision-makers and budget holders worry about governance and up-front technology costs.

The differences this time are the speed and scale of change. The time between gen AI capabilities being a competitive advantage and becoming a competitive necessity is dramatically shorter than it was in earlier technological transitions. Organizations that master the art of fast adoption will determine the new rules of their industries.

But how do leaders accelerate learning across an entire organization without sacrificing quality or creating chaos? How do they ensure that the enterprise gets the most value out of dispersed pockets of innovation? We discuss four mindsets and practices that can help.

Nurture what’s already growing

In the book The Gardener and the Carpenter: What the New Science of Child Development Tells Us About the Relationship Between Parents and Children (Farrar, Straus and Giroux/Macmillan Publishers, 2016), developmental psychologist Alison Gopnik argues that parents should allow children to develop according to their natural tendencies rather than predetermined constructs. This concept, which she calls a “gardener’s mindset,” is as relevant to organizational leaders as to parents: Nurture the growth that you see. The most successful managers focus on identifying the sprouts—employees, teams, or departments that are experimenting with new technologies and showing promising early results. They ask, “Where is innovation already happening? Who is solving problems in surprisingly effective ways?”

Most organizations, however, favor the “carpenter’s mindset”: meticulously planning every detail of technological transformation from the top down. This approach can’t keep pace with the current rate of change. Leaders who try to specify precisely how AI should be implemented across their organizations often find themselves building yesterday’s solutions for tomorrow’s problems.

Consider the experience of an Asian financial-services company that found its teams informally using AI to streamline application development. Managers embraced the innovation, creating a common data layer that allowed teams to automate time-consuming steps, such as data labeling, which cut AI application development times in half.

We have seen many similar examples, such as customer service teams quietly using AI chatbots to draft responses, often dramatically reducing their response times. Some management teams, worried about security or governance, shut down such experiments, while others study what makes them successful, refine the approaches, and help scale them. Recognizing and nurturing what’s already growing is more likely to advance innovation than trying to plant seeds based on theories. But applying a gardener’s mindset requires leaders to spend more time observing patterns and less time creating rigid plans. It means accepting that the most transformative ideas often emerge from unexpected places within the organization.

Create incentives for adoption

Everyone knows how difficult it is to change established work habits and learn new tools. The middle layer of most organizations—the managers and senior practitioners who set the cultural tone—is often the most resistant to change because of rational self-interest. They’re busy, their current methods work reasonably well, and the learning curve for new technologies can feel daunting.

Both financial and social incentives are essential to encourage meaningful adoption. But the most effective rewards focus on learning rather than just usage. Instead of offering bonuses for implementing AI, successful organizations reward employees for demonstrating new competencies, sharing insights with colleagues, and helping others navigate the learning curve. Social recognition often proves more powerful than financial rewards do. When respected team leaders share their AI learning journeys and publicly acknowledge that they’re still learning, it reduces the psychological barriers for everyone else.

Numerous large organizations (including McKinsey) run innovation competitions in which colleagues collaborate with diverse groups of peers and submit ideas. Teams that advance to later rounds may receive increased resources, expert support, and leadership exposure. The best companies offer such incentives not only during annual events but every day. One technology executive says that innovation rituals, including regular innovation days during which “teams explore interests and uncover ideas that may not be road-mapped yet,” are pervasive at their company. These sessions often yield unexpected discoveries that lead the organization to reprioritize its next wave of projects.

Foster rapid learning

Successful organizations don’t just experiment more than their peers; they experiment better. They borrow principles from the rigorous world of A/B testing and apply them to organizational innovation:

• Start with clear hypotheses. Instead of vague goals, such as “improve productivity with AI,” successful teams begin with specific, testable predictions—for example, “We believe that using AI to automate our monthly reporting process will reduce the time spent by 50 percent while maintaining accuracy above 95 percent.” But new ideas are only as good as their underlying assumptions, and all too often, teams don’t identify those assumptions or test them rigorously enough.
• Design for learning, not just success. Pilot programs often have successful outcomes as their only objectives, so they rarely generate valuable insights about what actually works. Better experiments are designed to fail fast and document approaches that didn’t work. And they span functions. For example, an R&D team can gain customer insights from salespeople early in its process. Such experiments also include control groups, where possible, and measure leading indicators, not just lagging outcomes. Some companies start from what a successful outcome would be, then identify the assumptions that would underpin it. An agricultural biotech company used this assumption-led approach to identify the best path to launch a new product and shortened the time to launch by 30 percent.
• Embrace the power of small samples. Organizations don’t need massive rollouts to generate meaningful insights. Some of the most valuable organizational experiments involve five to ten people over two to four weeks. The goal is rapid iteration, not statistical significance.
• Document the why behind results. Whether an experiment succeeds or fails, the most critical question isn’t “What happened?” but “Why did it happen?” Teams that systematically capture these insights build institutional knowledge that accelerates future innovation.

Amazon’s early attempts at video streaming exemplify these principles in action. Prime Video initially underperformed, but instead of scrapping the idea, Amazon asked why users weren’t engaging. The company found that customers didn’t see stand-alone value in the service and were drawn more to platforms with exclusive content. In response, Amazon bundled Prime Video into the broader Prime membership to increase perceived value and invested heavily in original content. This shift turned a struggling pilot into a key driver of Prime subscriptions and brand loyalty.

Maintain high standards for praise

In the enthusiasm to encourage innovation, business leaders often fall into the trap of celebrating everything equally. When every AI experiment receives hyperbolic praise and progress reports are allowed to exaggerate results in search of a bigger budget, the truly breakthrough ideas get lost in the noise. The most innovative organizations distinguish between interesting experiments (those worth trying) and game-changing innovations (those worth scaling). They reward honest reporting of failures as much as they celebrate successes.

This doesn’t mean being discouraging; it means being purposeful. When praise is selective and specific, it carries more weight. When leaders articulate exactly why a particular approach represents a breakthrough, teams understand what excellence looks like.

Organizations can transform their innovation culture simply by changing how they discuss pilot programs. Instead of asking, “How is the AI project going?” they ask, “What did you learn that surprised you?” Instead of celebrating that someone used AI, they celebrate the specific insights on better ways of working that emerged from its use.

A group CEO of a conglomerate, for example, has encouraged broad ownership of projects and a focus on tangible results. They asked 100 business leaders if they would each sponsor an AI project with specific targets for revenue increase, cost reduction, or customer satisfaction improvement. That target had to be reflected in the budget for the following year or the year after.

Organizations that master these principles don’t just adopt new technologies faster; they also develop a competitive advantage that compounds over time. Each successful experiment builds organizational confidence. Each well-documented failure prevents others from repeating the same mistakes. Each gardener-minded leader creates space for more innovation to bloom.

The future isn’t just unevenly distributed—it’s constantly being redistributed. Learning organizations are reaping the benefits of spotting innovation early, nurturing it carefully, and scaling it wisely.