Exploring generative AI effects on Korean civil servants’ performance

This study explores the factors influencing civil servants’ perceptions of generative AI-assisted work performance, focusing on extrinsic and intrinsic motivations and the moderating role of organizational power, as measured by government level and organizational rank. Artificial Intelligence (AI) has emerged as one of the most influential technologies, driving transformation in both government and society. Generative AI, as a subfield of AI, has gained recognition for its capabilities to automate routine tasks, enhance decision-making, and improve work productivity. However, little is known about who perceives the positive effects of these tools on performance and why. Drawing on the notion that motivations affect work performance and the reinforcement politics model, this study analyzes survey data from 1,608 Korean civil servants collected by the Korean Institute of Public Administration (KIPA) in April 2023, when generative AI was still in its early stages of development. The survey targeted civil servants from central, provincial, and local governments to gauge their perspectives on generative AI-assisted work performance. This study employed partial least squares structural equation modeling (PLS-SEM), an appropriate method for predictive modeling and identifying key driving factors in complex relationships to investigate the connection between two types of motivation and perceived work performance. The findings reveal that extrinsic and intrinsic motivations positively influence perceived work performance, with intrinsic motivation having a more substantial effect. Also, extrinsic motivation significantly enhances intrinsic motivation, highlighting the dynamic interaction between the two constructs. However, the moderating effects of government level and organizational rank on the relationship between motivations and perceived work performance were not statistically significant. These results underscore the critical role of individual motivation in shaping perceptions of generative AI tools while suggesting that organizational power may play a less significant role in this context than previously anticipated. Theoretically, this study provides empirical evidence supporting the idea that motivations affect work performance by demonstrating the effectiveness of generative AI tools in public administration while challenging the reinforcement politics model. It challenges the foundational assumption of the model that organizational power makes a significant impact on the adoption and effectiveness of generative AI tools.