Managers who routinely hand off thinking tasks to generative AI tools risk losing the practical wisdom they would normally build through direct experience, according to research from the University of Bath. The study, published in the Academy of Management Review, warns that this epistemic de-skilling chips away at the judgment, contextual understanding, and moral insight that come from solving real problems in real time.
Researchers from Bath, Ohio, Lausanne, and Cardiff universities examined how tools like ChatGPT affect what they call "managerial phronesis" - the hands-on wisdom developed through reflection, trial, and human interaction. When managers outsource idea generation and problem-solving to AI, they skip the very processes that forge this capability.
What gets lost when AI does the thinking
Professor Dirk Lindebaum, of Bath University's School of Management, said the shift creates "a significant risk" for organisations. "As managers increasingly outsource their thinking to Gen-AI for idea generation, or when a practical problem arises at work, they may rely less on their own judgment," he said. "Over time, this could reduce their ability to learn from experience, think critically, and to anticipate what kinds of actions are needed now to meet future goals."
The danger runs deeper than simple skill decay. Lindebaum described managers who stop asking important questions, stop seeking different perspectives, and stop learning from real-world interactions. Instead of building a nuanced understanding of employees, customers, or organisational challenges, they lean on AI-generated answers that lack the context and moral weight complex decisions demand.
This pattern mirrors what researchers call epistemic de-skilling - the gradual erosion of knowledge and experience-based capabilities as thinking is offloaded to machines. For managers, the loss is not technical know-how but the harder-to-measure capacity to read a situation, weigh competing values, and act with judgment that no algorithm can fully replicate.
The upskilling counterpath
The study does not dismiss generative AI outright. Researchers found that managers can use these tools to sharpen their thinking - provided they treat AI as a sparring partner rather than an answer key. This approach, which the researchers term epistemic upskilling, demands more from managers, not less.
"Rather than accepting AI outputs at face value, managers can use them to challenge assumptions, explore alternative scenarios and test the reasoning behind their own decisions," Lindebaum said. Because AI systems often cannot explain why they produce particular answers, the gaps in those explanations can push people to think more deeply about their choices and the downstream consequences.
This path requires what Lindebaum calls "persistent effort" to fill the explanatory holes AI leaves behind. The outcome is most likely when managers know they will be held accountable for their decisions and will have to articulate their reasoning - a structural safeguard that organisations can build into roles and workflows.
Designing work for human judgment
The findings point toward a deliberate design challenge. "It is becoming increasingly clear that simply introducing AI tools will not automatically improve decision-making or organisational performance," Lindebaum said. "Instead, organisations need to carefully design roles, responsibilities and workflows to ensure employees continue developing the human skills that AI cannot replicate."
For management teams already integrating AI for Management into daily operations, the study suggests a clear line between augmentation and atrophy. The same tool that accelerates a decision can, over time, dull the instincts that make the decision worth trusting. Organisations pursuing AI for Executives & Strategy will need to build reflection checkpoints into their processes - moments where managers must explain not just what they decided, but how they arrived at it.
Why this matters for managers
The core finding is practical and immediate: every time you ask an AI to solve a problem you could have wrestled with yourself, you trade a short-term speed gain for a long-term capability loss. The fix is not to avoid AI. It is to use its outputs as prompts for your own thinking - to interrogate the answer, map the gaps in its logic, and test your own assumptions against what the machine produced. If your organisation does not already require you to explain the reasoning behind AI-informed decisions, building that discipline yourself is the surest way to keep your judgment sharp.
Your membership also unlocks:
