Artificial Ladies against corruption: searching for legitimacy at the Brazilian Supreme Audit Institution

Abstract

This study depicts the search for legitimacy by four information technology artifacts in helping auditors in the surveillance against fraud and corruption by the Brazilian Supreme Audit Institution (TCU). ALICE, ADELE, MONICA, and SOFIA are Artificial Intelligence (AI) systems proposed to aid auditing processes in the public sector. A web-based survey has been used to gather the responses from 60 auditors across Brazil and semi-structured interviews with the Chief Data Officer, three IT Developers and five TCU Audit Managers selected by purposive sampling. The research finds that the use of AI-based systems is low among auditors at the TCU due to the perceived limited benefit. While some respondents recognized the advantages of the AI-based systems, they are put off by the weak theorization and diffusion regarding the meaning and the use of AI-based systems within the organization; they showed a priority for using traditional auditing methods instead of digital innovation, restricting the potential of anticorruption control by technological artifacts.

Keywords

• Artificial Intelligence Systems
• Auditing Processes
• Institutional Theory
• Corruption