Initial Adoption of Generative Artificial Intelligence in Management Reports: An Exploratory Study from the Perspective of Technology Acceptance

Abstract

The use of artificial intelligence systems advances in corporate environments due to the need for agility in decision-making processes. In this scenario, language models stand out as support tools in data analysis and in the preparation of management reports. Despite this utility, there is a scarcity of evidence regarding the interaction of managers with automation technologies in the early stages of adoption. For this reason, this exploratory study investigates the effects of ChatGPT use on the facilitating dimension, the intensity of use, and the technical validity of internal documents. For this purpose, the research applies a quantitative approach through a survey with seventy-three leaders of innovation companies. Subsequently, the evaluation of the hypotheses utilizes the Structural Equation Modeling (PLS-SEM) method. Thus, the results indicate a positive effect of the technology on the facilitating dimension and on the technical validity of the information. In contrast, the automation of routines results in a decrease in the intensity of use of the report by the manager. Finally, the conclusions show that the delegation of tasks to the machine alters the levels of human involvement and demands new control strategies in organizations.