A post on a Korean online community described a team working late every night — except one junior employee who left on time. He had used the company’s AI tools to build an automation for complex tasks. When a senior colleague suggested sharing it with the team, he refused: sharing would only mean more work for him.
It isn’t just a Korean story. According to a Harvard Business Review article, a KPMG and University of Melbourne survey of more than 48,000 people found 57% of workers had hidden their AI use at work. In the author’s own study of about 600 U.S. workers, roughly a third had deliberately hidden AI knowledge — while 80% believed sharing would raise team productivity. They know sharing helps and hide anyway. That tension sits at the heart of AI adoption and company culture.
Seventy-six percent of workers already use AI (McKinsey, 2026), 70% of skills in most jobs are expected to change by 2030 (LinkedIn, 2025), and 40% of workers fear losing their jobs to AI (Mercer, 2026). Korean HR tech company CLAP frames the impact through the four sources of meaningful work identified by Rosso and colleagues (2010).
When AI takes over long-practiced tasks, expertise can suddenly feel ordinary. People worry that sharing AI methods makes them look less capable, and AI-assisted work gets dismissed as “what the computer did.” Yet freeing people from routine work can also elevate judgment and creativity.
When everyone races to finish their own tasks with AI, knowledge-sharing slips. If sharing simply adds to your workload, AI becomes a tool for every person for themselves. Used well, it clears admin time for conversations that matter.
If AI collects work data for control, employees see surveillance, trust and psychological safety fall, and knowledge hiding grows. Explaining what data is collected and why, and discussing it openly, produces the opposite effect.
When AI can replicate expertise, people question whether their contribution is special. Used responsibly, AI can also create new purpose — serving more people or tackling problems that were once out of reach.
Marketing teams adopt AI tools faster than most. For a colleague’s prompt library or automation workflow to become a team asset, sharing must not cost the person who shared. Recognize sharing as performance, and let saved time go to skill-building. As OpenAI’s analysis of 17 million enterprise messages found, the AI gap between companies is widening — and it widens fastest where knowledge flows. Treating your organization like a community points the same way.
They fear more work, looking less skilled, or having results dismissed as the AI’s work. In a global survey, 57% said they had hidden AI use at work.
Self, others, work context and spiritual life or purpose, as defined by Rosso and colleagues (2010). AI adoption affects all four.
Make one-on-ones and peer feedback routine, use AI for recognition rather than surveillance, and agree with employees on how AI-freed time will be used.
To apply what you just read to your own site, start with a free audit of where things are now.
A strategist replies within 24 hours on business days.
Changing a work arrangement is a change to employment conditions, not an operational decision. Here are the procedures and documents each arrangement requires, and the three checks teams most often skip.
LLM visibility depends on your whole digital footprint, not your website. Teams are still built by channel. Mordy Oberstein's uncomfortable advice: do not reach for an action plan yet.
The welcome kit is ready but account setup gets missed. The problem is not a missing programme — it is that what, when and who were never assigned.