The Effects of the Use of Artificial Intelligence and Task Complexity on Auditor Liability

• In Event: All About Audit Evidence: Client Power, Auditor Empowerment, and Letting AI Have Some Power

Abstract

The increasing prevalence of artificial intelligence (AI) in public accounting can improve audit quality. However, drawbacks such as difficulty analyzing non-standardized human language that is unique to specific contexts can create challenges for the application of AI in the audit setting. We extend prior literature on how new audit technologies affect auditor liability by examining the effects of auditors’ use of AI and audit task complexity on jurors’ negligence verdicts. We link theories of algorithm aversion and attribution to develop our hypotheses. Results support our expectations such that: (1) jurors are more likely to render a negligence verdict when AI performs an audit procedure than when human auditors do so; (2) this effect is more pronounced for complex tasks than simple tasks; and (3) this effect is mediated by perceived trustworthiness of the audit approach. We offer contributions for theory, the juror literature, and practice.

Authors

• Junnan Cui, St. Norbert College

• Jesse C Robertson, University of North Texas