Artificial intelligence is already making people better at parts of their jobs. In a study of 5,172 customer-support agents, access to a generative AI assistant increased productivity by approximately 15 percent, with some of the strongest benefits seen among less experienced workers (Brynjolfsson, Li, & Raymond, 2025). The technology appeared to help newer employees access practices associated with stronger performers and move more quickly along the experience curve. It is an encouraging finding for the future of work. But it also raises a question I have become increasingly interested in through my own research: if AI helps us perform like experienced professionals earlier, how are we developing the experience required to eventually lead?
For a long time, performance and development naturally occurred together. A young professional drafted the memo, struggled through the analysis, received criticism, revised the work, made smaller decisions, observed consequences, and gradually took on greater responsibility. Those repetitions did more than produce an output. They helped develop judgment, confidence, self-awareness, and the ability to work through ambiguity. Today, AI can help create a polished memo in a fraction of the time. That may be excellent for productivity. But another question remains: what did the person learn while producing it?
This distinction between performance and development is at the center of the leadership-readiness research I am currently developing with Dr. Jeffrey Bohler, Associate Dean of the College of Business at Auburn University at Montgomery. Our work examines how generative AI use may relate to the development of future leaders, particularly when it comes to capabilities that are much harder to observe than productivity: judgment under ambiguity, reflection, self-awareness, communication, emotional intelligence, and readiness to assume greater responsibility. The project remains research in progress, and the goal is not to assume that AI weakens leadership development, but to examine how different ways of using AI may support or interfere with the experiences through which leadership readiness develops.
Existing research suggests that the answer may depend less on whether AI is used and more on how it is used. Lee and colleagues (2025), studying 319 knowledge workers and 936 examples of real-world AI use, found that higher confidence in AI was associated with less self-reported critical thinking, while greater confidence in one’s own ability was associated with more critical thinking. Importantly, these findings were correlational rather than causal.
Other evidence points in a more encouraging direction. Essel and colleagues (2024) found that students using ChatGPT within structured, scaffolded classroom activities demonstrated improvements in critical, reflective, and creative thinking. Similarly, Yusuf and colleagues (2024) tested a structured approach for critically analyzing and synthesizing AI-generated information and found that these skills could be developed through deliberate practice. Together, these studies suggest something important: the variable may not simply be the tool; it may be the design of the task around it.
This becomes especially important for early-career professionals. If AI increasingly performs routine work, younger employees may receive fewer repetitions of the tasks through which professional judgment was traditionally built. Routine tasks can be automated, while difficult exceptions continue to reach the person. Efficiency improves, but some of the developmental experiences hidden inside ordinary work may also change.
At the same time, expectations of leaders are becoming more complex. Leaders today are expected to adapt, collaborate, exercise ethical judgment, communicate effectively, understand technology, manage people, and lead through uncertainty. One direction we are exploring in our research is whether leadership expectations have accumulated over time rather than simply replacing older expectations. If so, an interesting tension emerges: we may be asking the next generation to develop a broader set of leadership capabilities while simultaneously changing some of the workplace experiences through which earlier generations developed them. Our current work treats this as a question to investigate, not a conclusion already established.
The question, then, should not be whether we should use AI. That debate is quickly becoming outdated. The more useful question is how we design AI-enabled work so that efficiency does not come at the expense of development.
Organizations and educators can preserve opportunities for independent judgment, ask people to explain the reasoning behind AI-assisted work, teach structured approaches for evaluating AI-generated information, and assess judgment rather than simply the polish of the final product. These are also among the practical directions emerging from the broader work surrounding our research.
AI can help people produce better work today. The challenge ahead is making sure that the way we use it also helps people develop the judgment, reflection, and human capabilities we will expect from them when they are asked to lead tomorrow.
