The application of artificial intelligence (AI) in the medical industry is becoming increasingly widespread. Relying on its powerful machine-learning capabilities, it has gradually become an important auxiliary diagnostic device. At the same time, it gradually has a certain degree of autonomy. But this also leads to the problem of lack of transparency in algorithms. A critical ethical issue known as the “algorithmic black-box” problem has emerged. For the ethical challenges associated with the opacity of medical artificial intelligence algorithms, enabling artificial intelligence systems to have reasoning and moral judgment abilities, namely constructing medical artificial moral agents, has been proposed as a viable solution. This article analyzes the solutions of constructing medical artificial moral agents including “top-down” and “bottom-up” approaches. After that, the new hybrid ethical design approach is proposed, integrating the advantages of both top-down and bottom-up approaches.
Qinggui, H. ., Xiuqiong, T. ., Hui, C. ., & Jinsong, W. . . (2025). The ethical risks and solutions of the algorithm black box in medical artificial intelligence. Acta Bioethica, 31(1). Retrieved from https://actabioethica.uchile.cl/index.php/AB/article/view/78363
