Businesses are increasingly using artificial intelligence (AI) in accounting systems to reduce uncertainty and improve accuracy. However, algorithm aversion (Dietvorst et al., 2015) indicates that individuals often avoid information provided by automated systems as compared to that provided by humans. This paper is an exploratory step towards documenting an emotional response to AI. We experimentally investigate how disclosing the use of AI rather than human staff for estimating the fair value of an asset influences investment decisions through lower levels of emotional response, particularly in pleasantness and attentiveness. Consistent with algorithm aversion, we find that disclosing the use of AI to estimate the asset’s fair value reduces the effect of information valence on nonprofessional investor responses. Specifically, when a company’s AI usage is disclosed, investors make smaller additional investments when fair value information is positive and smaller investment withdrawals when fair value information is negative, as compared to when human staff usage is disclosed. Importantly, we also find that emotions mediate the effect of information source (AI versus human staff) and moderate the effect of information valence on investment decisions.

中文翻译：

---

算法厌恶、情绪和投资者反应：披露人工智能的使用是否会影响投资决策？


企业越来越多地在会计系统中使用人工智能 (AI)，以减少不确定性并提高准确性。然而，算法厌恶（Dietvorst et al., 2015）表明，与人类提供的信息相比，个人通常会回避自动化系统提供的信息。本文是记录对人工智能的情感反应的探索性一步。我们通过实验研究了披露使用人工智能而不是人类员工来估计资产公允价值如何通过较低水平的情绪反应（尤其是愉悦感和专注度）影响投资决策。与算法厌恶一致，我们发现披露使用人工智能来估计资产的公允价值会降低信息效价对非专业投资者反应的影响。具体来说，与披露员工使用情况相比，当公司的人工智能使用情况被披露时，当公允价值信息为正时，投资者追加投资的金额较小；当公允价值信息为负时，投资者撤资的金额也较少。重要的是，我们还发现情绪调节信息源（人工智能与人类员工）的影响，并调节信息效价对投资决策的影响。