Turn individual experimentation into organizational learning

AI adoption is rapidly increasing in organizations, but merely using the technology does not guarantee a transformative impact on work. A Gallup report reveals that while 65% of U.S. workers in AI-implemented organizations perceive some positive effect on their productivity, only 12% strongly believe AI has fundamentally changed work processes. This discrepancy highlights the next challenge for leaders: converting individual AI benefits into organizational learning and workflow improvements.

Many early AI gains have focused on enhancing individual efficiency by speeding up existing work. However, the real opportunity lies in leveraging what one employee learns from AI to transform how an entire team, department, or organization operates. For instance, two employees working side by side might experience vastly different outcomes with the same AI tool.

One might use it to draft an email, while another might discover a process redesign that saves substantial time and resources. If these insights remain isolated, organizations miss out on their full potential.

As AI becomes increasingly integrated into business operations, the technology's accessibility alone may not sustain competitive advantage. Instead, the key differentiator will be an organization's ability to transform employee insights into shared knowledge, restructured workflows, and improved ways of working.

Leaders play a crucial role in fostering this transformation. They can leverage employees' firsthand knowledge of work processes to uncover practical applications of AI. For example, at Walmart, logistics managers contributed to developing an AI process aimed at improving truck efficiency and reducing empty trailers—an innovation that originated from someone deeply familiar with the workflow.

Leaders should define worthwhile problems to solve, set boundaries for experimentation, and articulate what constitutes an improved outcome.

Cengage, for instance, introduced AI in phases, empowering "Change Champions" across the company to experiment with AI, identify meaningful use cases, and disseminate their findings to their teams. This shared-learning model built organizational knowledge, allowing teams to benefit from examples and practices rather than starting from scratch as more employees gained access to AI. As a result, nearly 90% of employees with AI access are actively using it today.

Encouraging experimentation and creating a culture of shared learning are essential for turning individual AI discoveries into organizational benefits. Employees should have the ability to share their experiments and learnings, providing other teams with a foundation for their own exploration. At Canva, a dedicated week was allocated for employees to learn and experiment with AI, resulting in the organization-wide dissemination of insights gained during that period.

Traditional training can establish a shared understanding of AI usage and responsible practices, but it cannot keep pace with the rapid evolution of AI technology. Organizations must facilitate employees' ability to share their learnings, turning promising ideas into starting points for others. At Cengage, Change Champions used their subject matter expertise to experiment with AI, then shared practical examples with their teams, demonstrating the value of this approach.

Managers play a pivotal role in bridging the gap between experimentation and organizational scale. By fostering psychological safety around experimentation, managers can boost employee AI readiness and perceived value, as evidenced by Microsoft's 2026 Work Trend Index. Senior leaders should emphasize the importance of AI experimentation while managers provide the time, support, and guidance needed for employees to test ideas, identify worthwhile problems, and ensure responsible use of AI.

Managers are uniquely positioned to distinguish between an intriguing experiment and a transformative workflow improvement, acting as a crucial link between individual innovation and organizational learning.