The rapid digitalization of organizational processes has intensified the integration of artificial intelligence into project management, positioning it as a key enabler of automation, predictive analytics, and operational efficiency. However, the growing reliance on AI-based decision support carries significant risks that can compromise project outcomes and the sustainability of intelligent adaptive systems. This article examines the risks associated with overreliance on AI in project management and discusses its implications for adaptive systems in production and organizational environments. Based on a survey of 100 IT project practitioners and qualitative content analysis of open-ended responses, the study empirically confirms four dominant risk areas: data security and privacy vulnerabilities, algorithmic errors related to data quality, unclear legal and ethical frameworks, and overreliance on AI-generated results. Despite the increasing maturity of AI tools, the results show that the majority of respondents (78%) believe that AI cannot replace project managers due to the irreplaceable role of human judgment, creativity, and interpersonal skills, capabilities that remain crucial in complex and adaptive management ecosystems. Rather than replacing human decision-makers, the results highlight the need for hybrid models of human-AI collaboration, robust risk mitigation strategies, and effective governance mechanisms to ensure that intelligent adaptive systems support and complement, rather than replace, human decision-making in project environments.