The knowledge of ethical AI decision-making: A behavioral economics perspective

Focusing on artificial intelligence (AI) ethics and behavioral economics, this paper demonstrates how integrating behavioral insights can inform ethical guidelines for AI-based systems. Recent developments conceptualize AI as a key to critical sectors such as finance, marketing, and healthcare, thereby prompting a widespread recognition of its potential to reinforce or mitigate societal biases. However, while AI ethics has emphasized technical concerns such as data fairness and algorithmic transparency, insights from behavioral economics remain largely untapped. To bridge this gap in the literature, the present study uses a three-stage methodology. First, a systematic literature review (SLR) is conducted to identify gaps in AI ethics frameworks and to establish how key behavioral economic principles can enhance transparent decision making. Second, in-depth interviews are performed to collect additional insights and to compare how the results compare with those obtained from the SLR. Third, a topic modeling-based model, latent Dirichlet allocation (LDA), supported by the Computer-Assisted Text Analysis (CATA) framework, is applied to the collected interview corpus in order to identify the main topics in the data. The results reveal that ethical AI is increasingly understood by professionals not only as a technical issue but also as a behavioral and organizational one. Specifically, we identify five main topics from the interview content, namely Data & transparency, Behavioral insights & design, Legal/ethical frameworks, Implementation challenges & infrastructure, and Civic/stakeholder engagement. The first three of these topics dominate in the current discourse on AI decision making. The results also reveal that cognitive biases, such as anchoring and confirmation bias, affect the data that feed AI algorithms and user interaction with AI-driven outputs. If these biases remain unaddressed, they can amplify ethical risks. However, awareness of such patterns can inform practical design and governance strategies so as to improve transparency, fairness, and user engagement. Finally, based on the findings, four research proposals are advanced to clarify how behavioral economics can be systematically integrated into ethical AI decision making.