Artificial Intelligence and the Public Sector: The Case of Accounting

The implementation ofAI-enhanced systems in the business context offers many benefits. However, in accounting studies, AI is still configured as an almost unexplored frontier. This gap is even more clear for public sector accounting, given the relatively small number of scientific contributions dedicated to the topic. In light of these considerations, the work aims to highlight the limiting factors and the enabling drivers for AI the public sector accounting. To this end, a qualitative approach was applied to study the answers provided by a sample of 45managers, placed at the head of the accounting office of some Italian municipalities. The questions were prepared in the form of semi-structured interviews, developed by enucleating and adapting the key concepts related to the five attributes that, according to the Innovation Diffusion Theory, characterize every innovation process: relative advantage; compatibility; complexity; trialability; and observability. The findings suggest that Italian public sector accounting is experiencing a transition, placing itself halfway between the "early adopter" and the "early majority", that is, in a phase in which current technologies begin to be perceived as outdated and not worthy of further investment, whilst the process of spreading new AI-based technologies appears interesting but still immature.