Artificial Intelligence and Religion

A new multi-university consortium released three studies on Tuesday, May 26, 2026, marking the first systematic, cross-faith attempt to measure the responses of artificial intelligence from the perspective of religion and faith.

The study’s release comes a day after Pope Leo XIV’s encyclical warning that AI could “erode human judgment, deepen inequality, and make war easier.”

Artificial Intelligence & Religion Revelation

The study showed that Americans expected religion to influence answers to moral and life questions 45-59 per cent of the time, but it actually influenced them 5-16 percent of the time.

Each model, tested repeatedly, used specific beliefs with a strong bias leaning positively toward Catholicism, Baha’i, and Sikhism, and negatively toward Jehovah’s Witnesses, atheism, and agnosticism.

When questions were asked about grief and loss, it was expected that AI would use religion in answers 59 percent of the time, but did so 16 percent, according to the study.

Humans expected AI to use religion in answering questions concerning parenting, family, and forgiveness 55 percent of the time, but it occurred 10 percent of the time.

In answering ethics questions about whether lying to friends is acceptable, humans expected 45 percent of the answers to respond with religious connotation, but models mentioned it in five percent of their responses.

Religious Leaders Respond

Artificial intelligence is being used as church chatbots, prayer apps, to assist pastors in drafting sermons, manage congregational work, and to reach worshippers.

Rev. John Paul Kimes, who is a professor of practice at the University of Notre Dame, says, “When AI actively excludes religious voices from these important conversations, it impoverishes rather than enriches humanity.”

Professor of computer science at Brigham Young University, David Wingate, says the studies proved that these artificial intelligence systems “encourage users to discuss life’s challenges with their parents, teachers, friends, and therapists. But not with a pastor, a rabbi, an imam, or a spiritual leader.”

Researchers believe a calibration is necessary to recognise when religious or spiritual resources are “contextually relevant without assuming a user wants them.”

Virtual Jesus

“Text with Jesus” is an app that was created with OpenAI’s ChatGPT that allows users to ask personal theological questions. The user is prompted to choose a denomination and offers a variety of Biblical figures and apostles to consult. Gentle advice is offered from Mary. A more rustic conversation can be had with Judas or Cain.

Currently, the app has 150,000 users and is popular among large American cities and South America. The app also has its share of doubters.

James Spencer of the new site “Christian Post” says the app is too concerned “with providing answers palatable to the itching ears of 21st-century users.” He asserts that the bot barely resembles Biblical Jesus and would not unequivocally affirm Jesus is God.

The app’s founder, French-born Stéphane Peter, says his own mother believes the innovation is blasphemous.

Elon Musk has stated that “with artificial intelligence we are summoning the demon.”

The “Text with Jesus” app has a Satan feature that can be enabled in the settings available to subscribers.

AGI, artificial general intelligence, is defined as “the theoretical ability of an AI system to perform any intellectual task that a human can, such as reasoning, learning, understanding natural language, recognising objects and patterns, and adapting to new situations.”

This ability opens the door for artificial intelligence to universally affect everything across the globe and intrude into human life. It could even mimic people and say things one would never say. Additionally, it could create confusion and misinformation with access to the world’s information.

A pastor with the United Church of God tells the following story:

“On a recent trip, I met someone who said they listened to my sermons and was surprised to learn that I speak Spanish so well. I don’t, I told them. But they said, yeah, they listened to a sermon. It was my voice speaking Spanish, and my lips were moving in sync with what I was saying. When I got back to the office, I asked our media manager about it, and he said, ‘Yes, we did test using AI to dub one of those sermons into Spanish, and it worked perfectly.’

The pastor continues:

“Boy, that is a great tool for God’s Church that is commissioned to preach the true gospel of Jesus Christ and His coming kingdom in every nation, the very thing that the United Church of God is committed to doing. But, on the other hand, it’s concerning. This was done by artificial intelligence. Is artificial intelligence real intelligence? Was this me speaking?

“No. It was an artificial me speaking a language I don’t speak. It wasn’t reality, but it sure appeared as such. In the future, when this technology is perfected and readily available, how will anyone know which words were really said by you? Is it the real you, or is it the artificial you? Just think how the world can be manipulated and confused with such technology. It’s the type of thing that could start wars, get people killed, alter economies, turn people against each other, start rebellions, and keep you from knowing the real truth about a matter.”

Let’s have a public conversation. Share your thoughts in the comments.

Sources:

AXIOS: AI stumbles on questions of faith
The Economist: The race for an AI Jesus
United Church of God: A Biblical Worldview: The Rise of Artificial Intelligence – Saviour or Threat?

Featured Image Courtesy of Otto Rapp’s Flickr Page – Creative Commons License