Can We Teach Morality to Machines?

Artificial Intelligence and Ethics in Debate

  • Lucas de Azeredo Crespi UFRGS
Keywords: Ethics; Artificial Intelligence (AI); Logic; Knowledge; Representation; Cognitive psychology

Abstract

This review presents, analyzes, and critically discusses the work "Can We Teach Morality to Machines? – Artificial Intelligence and Ethics in Debate" (original title: Podemos Ensinar Moralidade às Máquinas? – Inteligência Artificial e Ética em Debate), published in 2025 by Dialética and written by Flávia Braga de Azambuja, who holds a doctorate in philosophy from UFPel. Derived from the author's doctoral thesis, the text seeks to ascertain whether an Artificial Intelligence can make moral decisions, which would imply an understanding of concepts such as right and wrong, permissible and forbidden, and good and evil. To this purpose, Azambuja investigates the possibility of representing abstract moral knowledge through modal logic. The book stands out for its interdisciplinary nature, as it deals with philosophical, technological, and even psychological themes. The text is divided into five segments, each with a well-organized internal structure and, in many cases, self-sufficient: the first part introduces the field of Artificial Intelligence and concepts such as Machine Learning, Deep-learning, Connectionism, and related terms. Also in part one, Azambuja discusses the consequentialist and utilitarian moral philosophy of John Stuart Mill, as well as the social constructivist approach of Jesse Prinz. In the second part, the work discusses mental logic, mental models, and the phenomenon of "belief bias." Chapter three is dedicated to modal logic, a type of non-monotonic formal logic that helps capture relations of possibility, necessity, and belief. In the fourth part, the preceding reflections on logic are used to address three moral dilemmas involving issues of public health, individual freedom, and national security. The last part of the book seeks to answer the initial question, finalizing the argument built throughout the entire work. "Can We Teach Morality to Machines?" is commendable as an introductory and accessible work. However, due to structural and methodological reasons, there are some insufficiencies in the treatment given to the subject of morality, which loses space to logic and cognitive psychology. The work, therefore, focuses too much on theoretical philosophy to the detriment of practical philosophy, which results in a final answer that is not fully satisfactory in view of the question posed, considering the breadth of perspectives and the intense debate that is inseparable from moral philosophy.

References

ANSCOMBE, Gertrude Elizabeth Margaret. Modern moral philosophy. Philosophy, v. 33, n. 124, p. 1-19, 1958.

ASIMOV, Isaac. Eu, Robô. Tradução de Luiz Horácio da Matta. 1969. Disponível em: https://kbook.com.br/wpcontent/uploads/2016/07/eurobo.pdf. Acesso em: 12 de abril de 2026.

AZAMBUJA, Flávia Braga de. Podemos Ensinar Moralidade às Máquinas? Inteligência Artificial e Ética em Debate. São Paulo: Editora Dialética, 2025.

AZAMBUJA, Flávia Braga de. Sobre a viabilidade da representação de conhecimento moral em sistemas inteligentes: uma abordagem baseada em lógica nãomonotônica. Orientador: Juliano Santos do Carmo. 2024. 114 f. Tese (Doutorado em Filosofia) – Instituto de Filosofia, Sociologia e Política, Universidade Federal de Pelotas.

ELLISON, Harlan. Não tenho boca e preciso gritar. 1967. Disponível em: https://notamanuscrita.com/wp-content/uploads/2015/06/visto-ellison-precisogritar.p df. Acesso em: 12 de abril de 2026.

ENGELS, Friedrich; MARX, Karl. A ideologia alemã. São Paulo: Boitempo, 2007.

HERF, Jeffrey. O modernismo reacionário: tecnologia, cultura e política na República de Weimar e no Terceiro Reich. Tradução de Cláudio Frederico Ramos. São Paulo: Ensaio, 1993.

HUEMER, Michael. Ethical Intuitionism. [s.l]: Springer Nature BV, 2008.

HUME, David. Tratado Na Natureza Humana: Uma Tentativa De Introduzir O Método Experimental De Raciocínio Nos Assuntos Morais. [s.l.]: Editora Unesp, 2009.

JOHNSON-LAIRD, Philip Nicholas. Human and machine thinking. New York: Psychology Press, 2015.

KANT, Immanuel. Crítica da Razão Prática. Petrópolis: Vozes, 2016.

KANT, Immanuel. Fundamentação da Metafísica dos Costumes. Tradução de Paulo Quintela. Lisboa: Edições 70, 2009.

LEIBNIZ, Gottfried Wilheim. Ensaios de teodiceia. [s.l]: WMF Martins Fontes, 2023.

MACINTYRE, Alasdair. Depois da Virtude: Um Estudo em Teoria Moral. 2. ed. [s.l.]: Editora da Universidade do Sagrado Coração, 2001.

MILL, John Stuart. Lógica das ciências morais. [s.l.]: Editora Iluminuras, 2020.

O’NEIL, Cathy. Algoritmos de destruição em massa: Como o Big Data aumenta a desigualdade e ameaça a democracia. Tradução de Rafael Abraham. Santo André: Editora Rua do Sabão, 2021.

PRINZ, Jesse. Is morality innate? Em: Moral Psychology. The Evolution of Morality: Adaptations and Innateness. Cambridge MA: The MIT Press, 2007. v. 1, p. 367–406.

QUINE, Willard Van Orman. De um ponto de vista lógico - Nove ensaios lógicofilosóficos. 1. ed. [s.l.]: Editora Unesp, 2011.

ROSS, William David. The Right and the Good. Oxford: Oxford University Press, 2002.

SEARLE, John Rogers. Minds, brains, and programs. Behavioral and Brain Sciences, v. 3, n. 3, p. 417–424, set. 1980.

VOLTAIRE. Cândido, ou o otimismo. Tradução de Mário Laranjeira. [s.l]: Penguin-Companhia, 2012.

WOODS, Roger. The Conservative Revolution in the Weimar Republic. London: Palgrave Macmillan, 1996.

Published
2026-04-14
How to Cite
de Azeredo Crespi, L. (2026). Can We Teach Morality to Machines? Artificial Intelligence and Ethics in Debate. Virtualia Journal, 1, 244-261. https://doi.org/10.5281/zenodo.19579419