Key points
- Employees may hide knowledge when they fear that sharing it could make them easier to replace with AI.
- Leaders’ words can reduce uncertainty, address concerns, and show employees that they still matter.
- Leaders should explain how knowledge will be used, give credit, and involve employees in AI changes.
- Motivating language works only when backed by fair decisions, follow-up, training, and growth opportunities.
Suppose you ask one of your colleagues to describe a procedure that they have mastered after many years. Perhaps they respond to your question, yet only in part, omitting some key steps. Or maybe they say the file isn’t ready, promise to get in touch with you, or say they can’t quite recall how they solved the problem last time
This may not be because of laziness, inadequate documentation, or quiet quitting. As Catherine Connelly and her colleagues point out in a recent study, knowledge hidingmeans deliberately concealing or not providing information that has been asked for; it can be evasive, by giving an answer that proves unhelpful; it can involve acting as if one doesn’t know; or it can consist of offering a sensible explanation for why one is not sharing what one knows
The hiding of knowledge has always existed, but artificial intelligence has given employees a new reason for doing it
The fear behind the silence
When a business decides to introduce AI, the people in charge usually concentrate on the technical aspects: What can we automate? How much time will it save? Employees might be asking a different, more personal question: If I show everyone exactly how I do my job, will I then become easier to replace?
The fear is not unreasonable. AI can now perform jobs that previously required years of experience, and many workers find themselves with fewer options for alternative employment if their jobs disappear. Under such circumstances, a request to ‘share your best practices’ may come across not as a means of cooperation but rather as a way of saying ‘help us to record what makes you valuable’. Knowledge thus acts as a kind of job insurance
Research has recorded this kind of reaction. Arias-Pérez and Vélez-Jaramillo discovered that the technological disruption caused by AI and robotics was linked to evasive hiding, playing dumb, and rationalized hiding. In their study of 402 South Korean employees, Kim and Kim found that employees who worried AI might threaten their jobs were more likely to keep useful knowledge to themselves
The point is clear: People don’t simply share more because technology makes sharing easier; rather, when they feel threatened by it, they may guard the one asset they still control: what they know
A leader’s voice matters
This is where spoken communication can go surprisingly far, or at least farther than a slide presentation declaring that ‘people are our greatest asset’.Employees need to hear a leader explain what is happening, acknowledge what it could cost them, and show where they fit in the future
Motivating language is useful here because it goes beyond positive talk. It combines three kinds of leader speech:
- Direction-giving language reduces uncertainty by explaining what is changing, what is not, and what people should do next.
- Empathetic language recognizes the fear, frustration, and effort involved.
- Meaning-making language connects the change to a larger purpose and helps people see why their contribution still matters.
In research I conducted with Jacqueline and Milton Mayfield, motivating language from immediate supervisors was associated with employee voice through transparency, openness to feedback, and a stronger sense of purpose. Voice and knowledge hiding are not opposites; an employee can remain quiet without actively hiding anything. Still, both involve a basic social calculation: Is it safe and worthwhile to reveal what I know?
We should be precise about the evidence. Researchers have not yet established a direct causal link between motivating language and reduced knowledge hiding. But the pieces fit. Babič and her colleagues found that strong leader-member relationships and a shared desire to help the team reduced knowledge hiding. Spoken communication is one of the main ways leaders build or destroy those conditions
Artificial Intelligence Essential Reads
The Educational Artifact Is No Longer Enough
The Zen of the Closed Loop
What leaders can say
- Name the uncertainty. Avoid empty reassurance. A leader might say: ‘AI will change parts of these roles. We have not made every staffing decision, and I will not pretend otherwise. Here is what we know, what we are still deciding, and when you will hear from me again.’ Clarity does not mean having certainty. It means being honest about what is known and what remains unknown.
- Explain how employees’ knowledge will be used. If employees are being asked to document their work or train an AI system, explain how that information will be used, who will benefit, and how contributors will be recognized. Employees are more willing to share when doing so creates a path to greater responsibility rather than a path out of the company.
- Protect status as well as employment. Leaders might say: ‘If your approach becomes the team’s new standard, your contribution will be visible. I want you to help us design the new process, not simply hand it over’. Giving people credit, ownership, and involvement shows that their expertise becomes more valuable when it is shared.
- Make it a two-way conversation. Ask: ‘What about this change feels threatening? What would help you use AI with confidence? What are we not seeing from your side?’ Then listen, even when the answers are uncomfortable. A listening session without meaningful follow-up teaches employees that silence is safer next time.
Conclusion
Language used to motivate is no magic. Employees will believe their leaders’ actions rather than their words if the leaders praise openness but still reward internal competition, or if they promise development right before an unexpected layoff. Communication is effective when it makes the actual choices clear and is supported by fair decisions, training, credit, and opportunities for people to take on new work
This doesn’t mean words are unimportant. When there is a sense of uncertainty, employees look to a leader’s way of speaking to detect signs of threat, respect, and belonging. Although AI can produce a fluent announcement in a matter of seconds, only a human leader can support it, take responsibility for the consequences, and reassure others that sharing knowledge will not diminish their value
In the AI era, speaking face-to-face forms an essential part of an organization’s knowledge infrastructure. The more advanced the technology becomes, the more carefully leaders need to explain where people remain important and the more important it becomes to demonstrate that sharing their knowledge gives the organization a future worth joining
Arias-Pérez, J., & Vélez-Jaramillo, J. (2022). Understanding knowledge hiding under technological turbulence caused by artificial intelligence and robotics. Journal of Knowledge Management, 26(6), 1476–1491. https://doi.org/10.1108/JKM-01-2021-0058
Babič, K., Černe, M., Connelly, C. E., Dysvik, A., & Škerlavaj, M. (2019). Are we in this together? Knowledge hiding in teams, collective prosocial motivation and leader-member exchange. Journal of Knowledge Management, 23(8), 1502–1522. https://doi.org/10.1108/JKM-12-2018-0734
Connelly, C. E., Zweig, D., Webster, J., & Trougakos, J. P. (2012). Knowledge hiding in organizations. Journal of Organizational Behavior, 33(1), 64–88. https://doi.org/10.1002/job.737
Kim, B.-J., & Kim, M.-J. (2024). How artificial intelligence-induced job insecurity shapes knowledge dynamics: The mitigating role of artificial intelligence self-efficacy. Journal of Innovation & Knowledge, 9(4), Article 100590. https://doi.org/10.1016/j.jik.2024.100590
Mayfield, J., Mayfield, M., & Nguyen, C. N. (2024). Raise their voices: The link between motivating language and employee voice. International Journal of Strategic Communication, 18(1), 75–92. https://doi.org/10.1080/1553118X.2023.2265926
