The Impact of AI Risk Scores on Business Managers’ Ethical Decision-Making

Artificial intelligence (AI) presents significant benefits to the organizations and to the individuals, environments, and stakeholders they impact. However, AI systems can also pose a risk of harm. Development of an AI risk score representing the potential risk of the system may assist business managers with the ethical decision on whether to deploy an AI system. While the quantification of risks associated with AI has received attention from researchers, limited research exists analyzing summarized AI risk scores and their impact on decision-making in practice. Expanding on integrated and behavioral theories of ethical decision-making, this quantitative experimental study found that the presence of an AI risk score can reduce the likelihood of an unethical decision, and thus may positively influence business managers faced with an ethical decision. The study also explored the potential influence of the AI system’s use case when an AI risk score is present; however, no significant influence was identified in the scenarios tested. This study has implications for practice for organizations developing, deploying, and using AI systems.