As companies pour billions into generative AI tools, middle and senior managers continue to withhold critical knowledge, hampering collaboration and slowing progress, according to new findings from ESSEC Business School.
Human psychology, not technology, is the barrier
Anil Kshatriya, assistant professor of management at ESSEC, examined recent enterprise studies alongside behavioural psychology research. His work reveals that the primary obstacle to effective knowledge-sharing in modern workplaces is not flawed technology, but deep-seated human psychology.
Middle and senior managers often hoard knowledge out of fear of losing status and control, a phenomenon known as psychological ownership. Professor Kshatriya explains that psychological ownership occurs when people develop a strong emotional attachment to ideas or expertise they have created, turning knowledge into a personal asset tied to identity, influence, or job security.
“When individuals feel a strong sense of personal attachment to the knowledge they’ve developed, they may view it as a source of identity, influence or job security. This emotional stake can create resistance to sharing, regardless of how seamless or intelligent the technology,” says Professor Kshatriya.
His research, drawing on experimental methods and interdisciplinary insights from accounting, economics, and psychology, shows how this behaviour persists even in organisations equipped with cutting-edge AI platforms.
Smarter tools, stronger silos
Advanced generative AI systems rely on comprehensive internal inputs to analyse data and suggest innovations. Yet when employees hesitate to contribute, these tools produce limited results.
The consequence, according to Kshatriya, is stark: organisations end up creating “smarter silos” instead of truly interconnected companies. Silos—long criticised in management literature—become reinforced as AI amplifies isolated pockets of information rather than bridging them.
Supporting evidence extends beyond his own research. A 2023 McKinsey report on organisational health found that high-performance cultures can intensify knowledge hoarding, as individuals guard internally generated solutions like prized possessions. Similarly, a 2024 Harvard Business Review study showed that employees in competitive environments are 30% less likely to share novel approaches voluntarily, even when prompted.
Changing what gets rewarded
To address this, Professor Kshatriya urges leaders to overhaul performance metrics that reward only those who originate ideas—the so-called “idea owners.” He advocates balancing recognition by celebrating employees who actively seek out, reuse, and build on others’ insights.
He recommends several practical interventions:
- Reward curiosity through recognition programmes that highlight questions asked and connections made.
- Prompt reflection via structured debriefs to capture tacit insights—those unspoken nuances often lost in daily operations.
- Apply social metadata such as digital badges or dashboards to show the real-world impact of shared knowledge.
- Use AI analytics to identify isolated teams or “silent nodes” in knowledge networks and flag groups overlooked in exchanges.
“When curiosity and collaboration are recognised, knowledge flows more freely, and AI can deliver far greater value,” says Kshatriya.
Aligning technology with behaviour
Professor Kshatriya stresses that technology alone cannot resolve the behavioural dynamics that inhibit sharing. “AI alone cannot resolve the deeper behavioural dynamics that inhibit sharing,” he says. “It is up to leaders to embed systems of recognition, safety and reciprocity that make contribution feel worthwhile and seeking feel smart. The organisations that succeed won’t be those with the most powerful tools but those that align the tools with norms that reward openness, humility and interdependence.”
This message aligns with his ongoing work at ESSEC, where he explores how management controls influence internal information flows. His earlier experiments, presented at conferences including the American Accounting Association meetings, demonstrate that symbolic rewards such as status recognition can reduce hoarding behaviour.
A critical moment for UK firms
For UK firms racing to integrate generative AI amid economic pressures, these insights arrive at a pivotal moment. While AI adoption is accelerating, addressing psychological barriers could unlock productivity gains estimated in trillions globally. Without cultural change, however, investments risk yielding diminishing returns.
Professor Kshatriya’s recommendations offer a roadmap. By fostering environments where sharing enhances rather than erodes status, companies can transform AI from a silo-strengthener into a true collaborator. As generative AI evolves, the human element remains the decisive factor in harnessing its potential for innovation and growth.