Wayward AI priest defrocked by online Catholic theology resource

Online confessions and the go-ahead to bless a newborn with Gatorade instead of holy water: A rogue clergybot has been axed after taking too much liberty with its priestly powers. Damian Klamka/ZUMA Wire/dpa

A popular US-based Catholic publisher has ditched an AI-based "priest" it debuted on its website late in April, after the rogue chatbot was found to be giving unorthodox replies to readers’ questions.

The AI "theologian" reportedly mixed up "commute" with "communion", performed unapproved online confessions and gave the go-ahead for baptisms to be carried out using Gatorade.

Catholic Answers announced that its new "Father Justin" chatbot is now to be replaced, first by a character named "Justin" and later by another as-yet-unnamed "lay" AI bot.

Catholic Answers, which publishes books and podcasts, describes itself as "a media ministry that serves Christ by explaining and defending the Catholic faith."

The publisher denied that Father Justin’s ill-fated and short-lived ministry ended in a laicization (the formal removal of a priest's authority), and said it wanted to continue using AI to help the organisation’s mission of giving "sound answers to questions about the Catholic faith."

Part of the problem with Father Justin, it seems, was that people were asking it to do things it was not "trained" for and which can only be done in real life by a priest, such as hearing confessions and being absolved from sins.

"We didn’t anticipate someone would seek sacramental absolution from a computer graphic," wrote Christopher Check, president of Catholic Answers.

And while recent advances in AI have fuelled concerns about job losses across a swathe of industries, Catholic Answers said its use of AI is intended to supplement, not replace, the work done by its human apologists, who it said are sometimes inundated with queries.

The Catholic Answers clergybot follows the setting-up last year of Magisterium.com, an AI-based effort backed by the Pontifical Oriental Institute (PIO) which aims to make Catholic Church teaching "accessible like never before." That bot, however, was quickly accused of "rookie errors."

Catholic Answers' defrocking of its AI clergyman came as the Italian government and The Holy See announced that Pope Francis would participate in a session on AI at the upcoming Group of 7 (G7) meeting of leading democratic industrialized countries.

"I am convinced that the Pope's presence will provide a decisive contribution to defining a regulatory, ethical and cultural framework for artificial intelligence," said Italian Prime Minister Georgia Meloni.