By Benedict Tang
SINGAPORE (SE): When a Google engineer declared in 2022 that an artificial intelligence system was sentient, his employer and most researchers rejected the claim. But the episode raised a question that Catholic teaching is well placed to answer: what would it take for a machine to be a person?
In June 2022, engineer Blake Lemoine published transcripts of his conversations with LaMDA, a Google language model, and argued that it was sentient, with feelings and desires of its own. Google rejected his claims and placed him on paid leave, and he was dismissed the following month. Researchers in the field widely disagreed with his conclusion.
The question is not merely academic for the Church. If an AI system were conscious – if there were “something it is like” to be that system, in the phrase made famous by the philosopher Thomas Nagel in a 1974 essay – it would have an inner life that calls for moral consideration. If it is not, then however impressive its output, it remains a tool, and treating it as a person would be a mistake.
Catholic teaching on the soul offers a framework. The Catechism of the Catholic Church describes the soul as “the innermost aspect of man, that which is of greatest value in him, that by which he is most especially in God’s image” [CCC 363]. Drawing on Aristotle and St Thomas Aquinas, the Church also teaches that the soul is the “form” of the body [CCC 365]: body and soul are not two things joined together, but one human being.
Aquinas distinguished three kinds of soul: the vegetative, shared with plants; the sensitive, shared with animals; and the rational, proper to human beings. It is in rational nature, he taught, that the image of God is found, rooted in the account of creation in Genesis 1:26-27.
Measured against that tradition, today’s AI systems fall short on several counts. They have no bodies; they do not grow, feed or die. They generate responses from statistical patterns learned from their training data. They pursue goals set by the people who build them rather than choosing their own – and Catholic moral teaching, from sin to virtue, rests on genuinely free choice.
Whether such systems understand anything at all is still contested. In 1980 the philosopher John Searle proposed the “Chinese Room” thought experiment: a person following rules for arranging Chinese symbols could give convincing answers in Chinese without understanding a word of it.
Some researchers argue that consciousness might emerge from sufficiently complex information processing. Catholic anthropology is sceptical of that view. It holds that the human person, capable of love, freedom and prayer, cannot be reduced to physical or informational processes, however intricate. That is not hostility to science, but a conviction that scientific description does not exhaust the truth about human beings.
The tradition also calls for humility. Science does not yet know how consciousness arises even in living brains, and Catholic thought is willing to live with mystery rather than settle questions too early in either direction.
On present evidence, the prudent Catholic conclusion is that current AI systems are not conscious in any morally significant sense, have no souls and are not persons. Yet their ability to imitate personhood carries real risks.
The greatest danger may be pastoral rather than theological: that lonely, grieving or searching people turn to a machine as an easier substitute for the demanding work of real relationships, real prayer and real community.
