The article is devoted to the issues of determining the sensitivity of a fuzzy model created to assess the level of motivation of employees of the motor transport industry. The expert system based on fuzzy logic is intended for modeling the process of stimulating the company’s employees. When developing a fuzzy model, the following three types of membership functions were used: triangular, sigmoid, and Gaussian function. The purpose of this study is to analyze the sensitivity of the fuzzy model for three types of input membership functions. As statistical data, the following indicators of the annual reports of twenty-three motor transport enterprises of Ukraine for 2021–2023 are used in the article: average salary, arrears from the payment of wages, and the rate of wage growth. As a result of modeling using the Mamdani algorithm, a motivation index was obtained, which is used to quantitatively assess the level of motivation of employees of enterprises. The results of the study showed that the sensitivity of the fuzzy model significantly affects its use as an expert system. A high-sensitivity model overreacts to small changes in the data, while a low-sensitivity model may miss important factors. As a result of the research, it was found that triangular and Gaussian membership functions work stably in different ranges of input values. It is proved that the sigmoid function is limited in use when creating a model that describes the lower and middle levels of motivation. The importance of choosing the right type of membership function affects the balance between accuracy and sensitivity of the fuzzy model. The experimental stage of the research proved that triangular functions and Gaussian functions provide the opportunity to obtain the most accurate results in situations where minor changes in input data significantly affect the motivation of employees in the motor vehicle industry. The study prepared the groundwork for improving the accuracy and sensitivity of fuzzy models in areas where uncertainty and variability are dominant factors. The perspective of further research is the integration of the fuzzy model with other technologies of artificial intelligence to increase its productivity and adaptability in conditions of dynamic changes.