Embedding AI in the workplace: more about HR than IT

Adopting AI tools in the workplace is not simply a matter of rolling out new technology. It also requires a commitment to safeguarding employee well-being, maintaining motivation and encouraging active engagement. In an article analysing 74 publications on the topic, five researchers – including NEOMA’s Shota Kawasaki – show that HR teams play a crucial role in ensuring that employees experience AI adoption positively.

Should we see AI first and foremost as a driver of business performance or as a threat to jobs? Does it make employees' day-to-day work easier by eliminating repetitive tasks or does it lead to cognitive overload? Does it generate stress and insecurity or does it open up new opportunities for employees to develop their skills?

These questions have been the focus of (often one-sided) debate for many years, and the researchers’ aim is not to advance their own viewpoint. Instead, they seek to highlight the ambivalent nature of AI, which may be perceived as either an asset or a threat depending on the context. 

AI implementation for employees—blessing or curse?

On the other hand, the research team draws attention to an issue that has largely gone under the radar: the way employees perceive AI depends primarily on how HR prepares, supports and facilitates its adoption. This matters far more than the tool’s power or the expertise of the IT team installing it.

The research is based on an analysis of 74 articles published between 2005 and 2024 in leading international journals. It applies to a wide range of AI technologies, including robots and robotic assistants, algorithmic management, chatbots, virtual assistants, AI-based job interviews, and facial recognition systems.

All these AI applications are designed to cut costs, enhance efficiency, improve quality, provide flexibility and deliver greater productivity. But they can also generate distrust, anxiety, loss of self-esteem, isolation, fatigue and sleep disorders among employees. Still, no company can operate successfully with state-of-the-art AI systems if its people are struggling. Well-being at work isn’t a perk; it’s a must.

The three facets of well-being: psychological, social and physical

The researchers identify three dimensions of well-being: psychological, defined as a subjective sense of comfort and a positive emotional state; social, based on workplace relationships and a sense of fairness; and physical, reflected in individual perceptions and physiological indicators of health or illness. 

AI influences all three dimensions of well-being because it imposes new demands that employees may not always be able to meet. The key to bridging this gap lies in HR practices tailored to this new reality.

Drawing on their analysis of 74 publications, the researchers were able to examine these practices in detail, organising their findings around three complementary areas: developing a digital culture and AI skills; boosting employee motivation by creating a climate of trust; and making employees active participants in the transformations driven by AI rather than passive users.

Developing AI skills for greater engagement

HR initiatives focusing on skills development foster stronger employee engagement while lowering stress and anxiety. The authors cite training as an example, with personalised learning paths designed to meet the needs of different employee profiles: which tasks can an AI system take on? What makes it so effective? How does it generate its answers? 

These training efforts will be even more effective if they are designed with the help of AI. Similarly, internal recruitment and candidate-selection processes can be improved with AI to pinpoint the people best suited to using it.

The authors also urge HR departments to put the tools in employees’ hands as early as possible so they can try them out and experiment with them. In addition, they recommend holding workshops and distributing tutorials and informational brochures.

Communication and transparency to counter the “black‑box” effect

The researchers recommend several strategies for boosting employee motivation around AI, including transparent communication, clear explanations of how algorithms work and fair AI-based performance evaluation systems that generate constructive feedback. Together, these measures help prevent AI from being perceived as a mysterious, impersonal and opaque “black box”.

Similarly, employees need reassurance that their jobs are not at risk. The authors stress that managerial culture plays a decisive role: if it is authoritarian or driven by internal competition, the introduction of AI will trigger insecurity, anxiety and defensive behaviour. Conversely, a management culture that supports employees, reassures them and fosters continuous learning will help mitigate these negative effects.

Redesigning jobs in partnership with employees

The article recommends involving employees in the transformation process to enhance their ability to act rather than restrict it. Since AI inevitably changes the content of existing jobs, this transition should be planned with employees themselves so they are actively involved and can move towards higher-value tasks. Another idea is to integrate “human” feedback processes into the tasks performed by AI to signal to employees that they still retain a degree of control over their work. 

Drawing on a wealth of examples, the article demonstrates that integrating AI into the workplace necessarily requires the involvement of HR decision‑makers. These leaders are not encouraged to pick and choose freely among the suggested courses of action, but to design comprehensive, coherent support mechanisms. This step is essential for harnessing the immense performance potential that AI offers.

Find out more

Sitong Yu, Shota Kawasaki, Federico Magni, Kang Yang Trevor Yu and Vijayan P. Munusamy, Optimizing AI adoption through HR practices: Systematic review and theoretical framework for improved employee well-being, Human Resource Management Review, September 2026. https://doi.org/10.1016/j.hrmr.2026.101152.