An empirical investigation of the expectations of AI's ethical decisions

2023 Fiscal Year Final Research Report

• Project/Area Number: 22K20320
• Research Category: Grant-in-Aid for Research Activity Start-up
• Allocation Type: Multi-year Fund
• Review Section: 0110:Psychology and related fields
• Research Institution: Waseda University (2023); Saitama Prefectural University (2022)
• Principal Investigator: Tanibe Tetsushi 早稲田大学, 文学学術院, 講師(テニュアトラック) (20964480)
• Project Period (FY): 2022-08-31 – 2024-03-31
• Keywords: 人工知能 / 自動運転 / 規範 / 公正感

Outline of Final Research Achievements

This study quantitatively examined the responses of the general public, who are not AI experts, to artificial intelligence (AI) judgments related to ethical issues. Study 1 investigated the evaluation of self-driving car behavior in the presence of contradictory norms related to car speed (compliance with legal speed and matching to the surrounding flow (actual speed)). The results showed a tendency to judge that compliance with legal speed is preferable in self-driving compared to driving by humans. Study 2 showed that when AI judgments were biased by gender, the negative impression may be alleviated compared to when humans make similar judgments. However, the effect sizes were small in both studies, and the robustness of the results remains to be examined.

Free Research Field

社会心理学

Academic Significance and Societal Importance of the Research Achievements

本研究の成果は、AIの開発や社会実装を社会的に受容される形で進めていく際に参照できる経験的な知見を提供する点で社会的意義があると考えられる。人間とAIでは期待される判断内容が必ずしも同じではないことや、同じ判断をしても異なった印象を与えうるという結果は、AIの判断にどの程度の柔軟性を持たせるかという設計上の問題や、AIに判断を任せてよい範囲を見極めるという運用上の問題を検討する際に役立つことが期待できる。また学術的な意義としては、倫理的判断の内容が同じでも判断主体が人か否かによって適切さの評価が変わることを定量的手法によって確認し、倫理に関して人々が持っている判断基準を解明することに貢献した。