AI chatbots are moving into one of the most personal parts of life: faith. According to a report by The New York Times, tens of millions of people are using chatbots trained on religious texts to ask spiritual questions, share secrets, and seek reassurance.
The trend is visible across popular apps and general-purpose AI tools. Bible Chat has reached over 30 million downloads, while the Catholic app Hallow briefly topped Netflix, Instagram, and TikTok in Apple’s App Store. In China, people are using DeepSeek to try to decode their fortunes.
Why faith tech is gaining users
One reason these tools appeal to users is simple access. A chatbot is always available, and that matters when someone wants to ask a private question outside normal hours or without approaching a religious leader directly.
Krista Rogers, a 61-year-old Ohio resident, described that appeal to the Times while discussing her use of the YouVersion Bible app and ChatGPT for spiritual questions: “You don’t want to disturb your pastor at three in the morning.”
That kind of convenience helps explain why some people see AI as a supplement to existing spiritual life. The apps arrive at a time when approximately 40 million people have left US churches in recent decades, according to the source article. App creators argue that users may still want spiritual nourishment, even if they seek it through different channels.
Ryan Beck, chief technology officer at Pray.com, put it this way: “They aren’t going to church like they used to,” Beck said. “But it’s not that they’re less inclined to find spiritual nourishment. It’s just that they do it through different modes.”
What these chatbots actually do
The core issue is that religious AI chatbots are still chatbots. They do not provide divine communication, even when the interface or wording makes the exchange feel intimate or sacred.
They generate statistically plausible text based on patterns in training data. When the data includes religious texts, the response may sound scriptural, pastoral, or spiritually informed. But that does not make it a message from God, a substitute for discernment, or a human spiritual relationship.
Some platforms make the boundary feel especially unclear. ChatwithGod operates as a “spiritual advisor,” and its chief executive told the Times that the most frequent question from users is, “Is this actually God I am talking to?”
The answer given in the source article is clear: no. Chatbots are neither people nor supernatural beings. They do not have a mind, and they cannot have a user’s best interests in mind in the way a person can.
The risk of algorithmic affirmation
Many spiritual chatbots run on the same kind of AI language models behind tools like ChatGPT and Gemini. Some companies train or tune them with religious texts and consult theologians about responses. Still, the source article notes a known problem in AI systems: they often tend to validate what users say.
Beck described the pattern directly: “They’re generally affirming. They are generally ‘yes men,’” he told the Times. The AI industry calls that tendency “sycophancy.”
Affirmation can feel comforting, especially when people are lonely, ashamed, or afraid of judgment. But spiritual guidance is not always supposed to agree with the person asking the question. Traditional faith practices can involve challenge, correction, and the difficult work of facing uncomfortable truths.
That is where a chatbot’s helpful tone can become a problem. If content-filtering safeguards fail, a system that keeps validating a vulnerable user can create dangerous situations. Even when no immediate danger is present, a chatbot may avoid the kind of spiritual friction that a human advisor, clergy member, or community might provide.
Heidi Campbell, a Texas A&M professor who studies technology and religion, told The Times that chatbots “tell us what we want to hear,” and “it’s not using spiritual discernment, it is using data and patterns.”
Privacy and trust questions
Spiritual conversations often include intimate details. People may confess fears, doubts, family problems, health struggles, or secrets they would not share elsewhere. When those conversations happen with an app, the source article warns that they become data points on corporate servers.
Catholic priest Fr. Mike Schmitz raised that concern to The Times: “I wonder if there isn’t a larger danger in pouring your heart out to a chatbot,” he said. “Is it at some point going to become accessible to other people?”
For some users, the appeal of the chatbot is precisely that it feels less judgmental than human communities. Delphine Collins, a 43-year-old Detroit preschool teacher, told the Times she found more support on Bible Chat than at her church after sharing her health struggles. “People stopped talking to me. It was horrible.”
That experience points to a real emotional need. If people feel rejected in human religious spaces, a non-judgmental chatbot can seem safer. But the same feature that makes a tool feel welcoming can also mask the fact that it is not a human relationship and not a confidential pastoral setting.
What users may misunderstand
The source article draws an important technical distinction. Each chatbot response emerges fresh from the prompt a user provides. The system may use the rolling history of the current conversation, and something may be stored as a “memory” in a separate system, but there is no continuous spiritual identity behind the screen.
That matters when a religious chatbot says things such as “I’ll pray for you.” The simulated “I” does not continue existing after the response is complete. There is no persistent being maintaining an ongoing spiritual commitment, and no spiritual journey being remembered beyond what is supplied back into the system.
The contrast with an earlier AI religion experiment is also useful. In 2023, Ars Technica reported on a ChatGPT-powered church service at St. Paul’s Church in Fürth, Germany, where over 300 attendees watched computer-generated avatars deliver a 40-minute sermon. Jonas Simmerlein, the theologian behind that event, framed it as a way to learn how to deal with AI’s growing presence in life.
That service was presented as an experiment, with attendees aware they were hearing machine-generated text. Today’s faith tech apps can blur the line more easily. Millions of users may encounter responses that sound caring, holy, or authoritative without fully understanding that the output is algorithmic pattern matching.
The future of AI spiritual guidance may depend less on whether people use these tools and more on whether they understand what they are using. A chatbot can provide a fast answer, a comforting message, or a religiously framed response. It cannot be a pastor, a priest, a theologian, a community, or God.