Cognitive biases in implementing intelligent decision-support systems in public administration: theoretical and empirical analysis

The limited consideration of cognitive biases in implementing intelligent decision support systems in Russian federal executive authorities has been studied. The main groups of cognitive distortions to which a person using artificial intelligence technologies is subject have been analyzed and described. The purpose of the study is to assess how psychological factors are integrated into digital transformation projects. The methodology includes a survey of 30 employees of 6 federal executive authorities. The results showed that 87% of respondents recognize the importance of cognitive risks, but only 20% fix them in regulations. The key barriers are staff resistance (53%) and lack of competence (50%). The hypothesis of systematic disregard of cognitive factors has been confirmed, which reduces the accuracy and credibility of the intelligent decision support systems. The practical significance of the results lies in formulating recommendations for implementing human-oriented design, cognitive resilience metrics, and mandatory training modules for civil servants. The conclusion has been made about the need to institutionalize psychological expertise as an integral part of the state strategy for artificial intelligence development. The prospects for further research include expanding the sample, factor analysis of the impact of user experience and role, as well as key performance indicators development for monitoring cognitive risks at all stages of the intelligent decision support systems lifecycle.