As companies increasingly use artificial intelligence (AI) to make decisions about hiring, promotions and pay raises, new research from the Chinese University of Hong Kong (CUHK) Business School suggests the technology may come with an overlooked human cost.
The study finds that employees respond substantially more negatively when important career decisions are made by AI systems instead of human managers.
The research, titled “AI in Human Resource Management: A Driver of Organisational Dehumanisation and Negative Employee Reactions,” was led by Professor Choi Sungwoo, Assistant Professor at the School of Hotel and Tourism Management of CUHK Business School, in collaboration with researchers from the Hong Kong Polytechnic University and the University of Macau.
Based on three online experiments involving nearly 700 participants, the study examined employees’ reactions to AI-driven promotion and performance-review decisions.
Across all experiments, participants consistently reported lower organisational commitment, stronger intentions to quit, and greater retaliatory feelings when AI was responsible for workplace decisions.
The researchers identify this as “organisational dehumanisation” — the feeling of being treated as a functional unit rather than as an individual with unique qualities and emotions.
According to the study, employees perceive AI systems as lacking empathy and failing to account for social and ethical nuances. Many also find AI decision-making processes opaque and difficult to understand, leaving them feeling powerless and excluded from the decision-making process.
“Putting it all together, loss of empathy, transparency, and control can leave people feeling objectified,” says Professor Choi. “When AI performs human resources operations, employee characteristics are seen as numbers. Therefore, employees would feel like they are not treated as humans.”
The study further identifies what researchers describe as a “cultural paradox” — companies with highly collaborative and people-oriented workplace cultures may experience stronger resistance when adopting AI in HR functions.
In such environments, employees tend to expect management to value cooperation, support and interpersonal relationships. The researchers note that AI-driven personnel decisions may therefore clash with employees’ expectations of a human-centred workplace.
The researchers stress that companies should not abandon AI altogether but instead adopt hybrid systems in which humans remain accountable for consequential decisions.
“This hybrid approach helps preserve the sense that consequential decisions about people are ultimately made by people,” Professor Choi says. “If AI serves only in an assistive capacity with limited input into the final decision, the dehumanisation effect should be substantially mitigated.”
