ARTIFICIAL INTELLIGENCE AS A MEMBER OF CRISIS MANAGEMENT LEADERSHIP – BENEFITS, LIMITATIONS, AND THE PSYCHOLOGICAL IMPLICATIONS OF HYBRID DECISION-MAKING

This article examines the integration of artificial intelligence as a constituent element within crisis management leadership structures, focusing on the technological, psychological, and ethical implications of hybrid decision-making systems. The study addresses a significant gap in scholarly attention regarding the psychological and ethical consequences of AI integration in emergency management, while acknowledging that technological dimensions have been relatively well-investigated. The primary objective is to present both the potential benefits and principal risks associated with incorporating AI into crisis response teams. The research highlights several functional advantages of AI integration, including enhanced predictive capabilities through real-time data synthesis, improved logistics optimization, and significant reduction of cognitive burden on human decision-makers. AI-supported predictive models can reduce median response times by up to 30% in countries employing machine learning-based systems, while optimization algorithms demonstrate 10–18% improvements in emergency unit response times compared to static distribution models. However, the study identifies substantial risks including automation bias, where human operators accept algorithmic recommendations uncritically, and the “black box” problem where AI decision-making processes lack transparency. Additional concerns include dependence on data quality, the erosion of individual and collective autonomy, information overload, and the emergence of accountability gaps in decision-making processes. The integration significantly impacts team dynamics, with potential effects on trust, autonomy, psychological safety, and communication patterns within crisis management teams. The research emphasizes the “responsibility gap” phenomenon, where attribution of accountability becomes ambiguous when decisions are influenced by algorithmic systems. The article concludes with foundational recommendations for safe and effective AI deployment, emphasizing the “human-in-the-loop” approach, ensuring system explainability, comprehensive training programs, and adherence to European legislative frameworks, particularly the Artificial Intelligence Act. The authors stress that successful integration requires combining technical precision with human responsibility, viewing AI as a complement rather than replacement for human judgment in crisis management contexts.