From sceptics to enthusiasts: how public auditors confront the AI era
Natalia Alonso-Morales, Alejandro Saez-Martín, Antonio López-Hernández, Dolores Genaro-MoyaPurpose
This study aims to examine auditors' readiness and attitudinal profiles toward for implementing artificial intelligence (AI) in public sector external auditing, comparing professionals working in institutions affiliated with the European Organisation of Supreme Audit Institutions (EUROSAI) and the European Organisation of Regional Audit Institutions (EURORAI).
Design/methodology/approach
Using survey data from 557 external public auditors, the study combines comparative tests, k-means clustering to identify readiness profiles, and multi-group partial least squares structural equation modeling to assess whether the determinants of willingness to implement AI differ across the two institutional settings.
Findings
The results suggest an early-stage adoption pattern in which auditors perceive strong AI benefits – particularly in automation, data analysis, and text processing – yet willingness to implement AI remains moderate. This gap is consistent with an audit context where adoption requires not only expected gains but also confidence in explainability and evidential traceability, alongside sufficient organisational support. The comparison between EUROSAI and EURORAI indicates selective rather than pervasive differences: willingness is more closely linked to enabling conditions in EURORAI, while effort-related perceptions play a more visible role in EUROSAI. Additionally, three readiness profiles (sceptical, moderate and enthusiastic) highlight substantial internal heterogeneity.
Originality/value
The study brings public audit institutions into the public management debate on AI-enabled digital transformation by integrating auditor heterogeneity with a comparative perspective. It shows that AI-acceptance mechanisms are selectively conditioned by institutional setting, underscoring the importance of considering both individual perceptions and contextual conditions when designing AI implementation strategies in public auditing.