The Attenuating Effect of Intelligent Agents and Agent Autonomy on Managers' Ability to Diffuse Responsibility for and Engage in Earnings Management

Abstract

SYNOPSISAdvances in IT suggest that computerized intelligent agents (IAs) may soon occupy many roles that presently employ human agents. A significant concern is the ethical conduct of those who use IAs, including their possible utilization by managers to engage in earnings management. We investigate how financial reporting decisions are affected when they are supported by the work of an IA versus a human agent, with varying autonomy. In an experiment with experienced managers, we vary agent type (human versus IA) and autonomy (more versus less), finding that managers engage in less aggressive financial reporting decisions with IAs than with human agents, and engage in less aggressive reporting decisions with less autonomous agents than with more autonomous agents. Managers' perception of control over their agent and ability to diffuse their own responsibility for financial reporting decisions explain the effect of agent type and autonomy on managers' financial reporting decisions.