Whether AI adoption challenges matter for public managers? The case of Polish cities

A growing body of literature shows that despite the significant benefits of artificial intelligence (AI), its adoption has many unknowns and challenges. However, theoretical studies dominate this topic. Completing the recent works, this article aims to identify challenges faced by public organizations when adopting AI based on the PRISMA Framework and an empirical assessment of these challenges in the opinion of public managers using survey research. The adopted research procedure is also an added value because it could be replicated in other context scenarios. To achieve this paper's aim, the Systematic Literature Review (SLR) and survey research among authorities in 414 Polish cities were carried out. As a result, a list of 15 challenges and preventive activities proposed by researchers to prevent these challenges have been identified. Empirical verification of identified challenges allows us to determine which of them limit AI adoption to the greatest extent in public managers' opinion. These include a lack of strategy or plans to initial adoption / continued usage of AI; no ensuring that AI is used in line with human values; employees' insufficient knowledge of how to use AI; insufficient AI policies, laws, and regulations; and different expectations of stakeholders and partners about AI. These findings could help practitioners to prioritize AI adoption activities and add value to digital government theory.