Moral Philosophy Coding: An Agent-Based Computational Experiment on AI Ethics

Abstract The development and application of artificial intelligence (AI) technology have raised ethical challenges such as technological unemployment while enhancing efficiency. This study presents a computational experiment-based ethical analysis method by encoding the ethical principles of utilitarianism and deontology into computable parameters through agent-based modeling (ABM), constructing an ethical evaluation framework regarding the diffusion and application of AI, and providing decision-making recommendations for technology governance that balance efficiency and equity after comparing the results of simulations based on different ethical principles. Preliminary results show that policy schemes that introduce compensation mechanisms based on the Kaldor-Hicks improvement and integrate utilitarian and deontological principles to some extent outperform those based on purely utilitarian or deontological approaches in terms of balancing efficiency gains with social equity. The model simplifies the complexity of reality while quantitatively demonstrating the real-world application of different ethical theories, promoting the advancement of responsible AI whilst providing methodological innovation for applied ethics research as well as practicing a path for the transdisciplinary integration of humanities, social sciences and computer science. Keywords: Applied Ethics; Agent-Based Model; Technological Unemployment; Utilitarianism; Deontology