Conference Agenda

• Objectives: The main objective of this study is to propose a new tool called GPT-QUAL, a multidimensional scale designed to evaluate user perceptions of the service quality of generative AI conversational agents, with a particular focus on ChatGPT.

• Theoretical Framework: GPT-QUAL goes beyond traditional service quality frameworks by incorporating additional dimensions such as personalization, privacy, and security, addressing the specific needs of AI-driven interactions. Its development follows Churchill’s (1979) construct development process.

• Methodology: The dimensions were identified and refined through a Delphi study involving AI experts and industry professionals. The psychometric properties of GPT-QUAL were validated using Structural Equation Modeling (SEM) based on a sample of 400 ChatGPT users.

• Results/Implications: The final scale consists of five dimensions: Efficiency, Assurance, Interface and Presentation, Personalization, and Privacy and Security. GPT-QUAL provides a valuable tool for researchers and industry professionals to better understand user perceptions, enabling improvements in service quality and enhancing user satisfaction with AI-based chatbots.

Objectives: This study examines how generative artificial intelligence (AI) tools influence employee performance in small and medium-sized enterprises (SMEs). The main objective is to understand how AI-specific job resources affect engagement and exhaustion, and how trust in these tools mediates these relationships.

Theoretical Framework: The Job Demands-Resources (JD-R) theory is used as the conceptual framework. This theory provides a foundation for exploring the impact of AI-specific job demands and resources on employee performance and well-being.

Methodology: The study is based on a survey conducted in November, with 465 responses from employees in the United Kingdom who use generative AI tools in their workplace. A structural equation model was used to analyze the relationships between job resources, demands, trust, engagement, and exhaustion.

Results/Implications: AI-specific job resources positively influence engagement, which, in turn, improves performance. Trust mediates the relationship between resources and engagement and helps reduce exhaustion, although it does not moderate the direct impact of demands on exhaustion. Employees in small enterprises (<50 employees) experience higher exhaustion levels than those in medium-sized organizations. To fully harness the potential of AI tools, organizations must build trust, enhance employee engagement, and address factors contributing to exhaustion.

Artificial intelligence (AI) is reshaping society, altering how workers perform and evaluate their job functions. However, the application of AI into hiring processes, such as in screening and selecting suitable job candidates, has provoked skepticism. In this study, we build and test a model that investigates whether feedback delivered by AI systems influences justice perceptions, and subsequently the reactions (endorsement and acceptance) of AI systems for hiring purposes. Using a 3 (feedback valence: positive vs. neutral vs. negative) × 2 (feedback recipient: self vs. others) between-subjects experimental design, we found evidence that negative feedback increases perceptions of procedural injustice, which undermines reactions of AI systems. Further analyses reveal that this relationship is stronger when negative feedback is directed towards the self instead of others. These findings provide pivotal implications for management theory and practice seeking to implement AI systems within workplace processes.

• Objetivos. El artículo analiza la relación entre la Inteligencia Artificial y su humanización en las organizaciones, explorando su relevancia científica, los principales temas de investigación y las tendencias actuales y futuras en la aplicación de tecnologías de IA humanizadas en el entorno organizacional.

• Marco teórico. Las organizaciones están en constante búsqueda de la innovación y optimización de sus procesos. La implementación de tecnologías con IA que emulen características humanas se presenta como una decisión estratégica que permite a las organizaciones mejorar la interacción entre humanos y dispositivos.

• Metodología. Se obtuvieron 171 artículos de Scopus tras una revisión sistemática de la literatura a través del modelo PRISMA. Se realiza un análisis de co-palabras y de acoplamiento bibliográfico.

• Resultados/implicaciones. Los resultados indican que la investigación sobre tecnología con IA humanizada es un campo emergente con un notable crecimiento desde 2019. Se identificaron tres bloques de estudio al respecto: gestión organizativa, interacción humano-robot, y aspectos psicológicos/sociales. Además, se distinguen cuatro tendencias clave de estudio: La humanización de la IA y la experiencia del usuario; la implementación de IA en las organizaciones; la ética, transparencia y confianza en la IA y por último, tendencias y futuro de la IA en el trabajo.