The rapid diffusion of Artificial Intelligence (AI) is transforming job characteristics, raising important questions about how to implement these technologies in organizations in ways that support employee well-being and performance. Drawing on the High-Involvement Management framework, this study examined employee-centered Artificial Intelligence implementation (ECAII) practices (defined as transparent communication, consultation, and training initiatives) as strategic levers to foster positive employee outcomes during Artificial Intelligence-driven transformations. Survey data were collected from 168 Italian white-collar employees who actively used Artificial Intelligence in their work. Structural equation modeling was employed to test direct and indirect relationships among employee-centered Artificial Intelligence implementation practices, work meaningfulness, job satisfaction, and job performance, as well as the moderating role of personal attitudes toward AI. Results showed that employee-centered Artificial Intelligence implementation practices had significant direct effects on both job satisfaction and performance, as well as indirect effects through work meaningfulness. Latent moderated mediation analyses further revealed that these indirect effects were stronger among employees with more positive attitudes toward Artificial Intelligence. Overall, the findings highlighted the importance of employee-centered strategies for enhancing meaningfulness and fostering positive outcomes during technological change. This study contributed to Human Resource Management (HRM) and meaningful work research by extending classic theoretical frameworks to Artificial Intelligence-enabled workplaces. Furthermore, from a practical perspective, our findings provided valuable guidance for organizations by highlighting the importance of transparent communication, employee involvement, and targeted training in reducing uncertainty and helping employees perceive their roles as relevant during the implementation of Artificial Intelligence.
Gemmano et al. (Sun,) studied this question.