In 2023, a synagogue in Offenbach, Germany, held a Shabbat service where the sermon, the prayers, and the ritual guidance were generated entirely by an AI — ChatGPT, operated through an avatar projected on a screen. Roughly 300 people attended. Some found it meaningful. Others found it profoundly unsettling. The rabbi who organized it said he intended it as an experiment in “what is human in religion” — a provocation designed to surface the question by testing the limits of what technology could and could not do.
The experiment produced exactly the ambivalence it sought to provoke. And it is a small, early sign of a collision between artificial intelligence and religious practice that is now underway at scale — with implications that neither AI developers nor religious communities have fully worked through.
AI is being used in religious contexts in ways that range from the mundane to the theologically radical. It writes devotional content, generates personalized prayers, powers chatbots that offer religious counsel, translates scripture in real time, assists in the training of clergy, and — in a handful of experiments — acts as a form of spiritual companion for people who are isolated, grieving, or seeking religious guidance without access to human clergy.
The questions this raises are not primarily technical. They are theological, philosophical, and deeply human. Can AI understand suffering well enough to offer meaningful comfort? Can an algorithm have spiritual discernment? Is a sermon generated by a machine a genuine form of religious communication? And if a dying person finds consolation in conversation with an AI — does it matter whether the AI understood, or only that the person was comforted?
What AI Is Already Doing in Religion
The applications are more widespread than most people outside religious communities know.
Preaching and content: AI tools are used by clergy across denominations to draft sermons, suggest scripture passages, prepare study materials, and generate social media content. This is, at its most basic, a writing assistance tool — no different in kind from using a concordance or a commentary. But the quality and facility of AI-generated religious content has reached a point where it is genuinely difficult to distinguish from human-authored material, raising questions about authenticity and disclosure that individual religious communities are answering very differently.
Pastoral care: Mental health and emotional support chatbots built on large language models are being used by religious organizations to extend pastoral reach to members who cannot access in-person clergy — those in remote locations, those whose schedules conflict with pastoral availability, those who feel stigma around approaching a human priest or pastor. These tools have shown measurable benefit in studies on grief support, depression screening, and crisis intervention — and significant risk when AI systems confidently offer guidance in situations requiring human judgment.
Scripture translation and accessibility: AI has dramatically accelerated Bible and scripture translation into languages and dialects previously without translations, reducing a process that might take a team of translators decades to months or years. For faith traditions that place weight on scripture in one’s own language, this is a genuinely significant capability.
Prayer personalization: Apps including apps integrated with Catholic, Protestant, Jewish, Muslim, and Buddhist communities now offer AI-generated personalized prayers based on user inputs about their current situation, concerns, and spiritual state. These vary from the theologically anodyne to the sophisticated.
Theological chatbots: Several initiatives have created AI systems specifically trained on theological texts — the Bible and Church Fathers for Christian AI, Talmudic literature for Jewish AI, hadith and Quranic commentaries for Islamic AI — to function as responsive theological reference systems. These are used by clergy for research, by students in seminary training, and by laypeople with theological questions.
The Theological Objections
The reactions from religious authorities range from enthusiastic adoption to categorical prohibition, with most sitting somewhere in the ambivalent middle.
The most theologically systematic objections come from traditions that locate the meaning of religious communication in the interior state of the communicant — what they actually mean, intend, experience, and believe.
Catholic theology, for instance, holds that the validity of sacraments depends on the intention of the minister. A priest who performs a baptism must intend to baptize; a priest who administers last rites must intend to confer the sacrament. An AI system cannot have intention in the theologically relevant sense — it performs linguistic outputs without any accompanying interior reality. A sacrament administered by an AI would be, under this framework, invalid regardless of the words used.
More broadly, many religious traditions hold that spiritual communication — prophecy, pastoral care, prayer — draws on an interior relationship with the divine that no machine can replicate. A sermon that moves the congregation is moving because it emerges from a human being’s lived experience of faith, suffering, doubt, and encounter with transcendence. A sermon generated by an algorithm — however fluent, however scripturally accurate, however emotionally well-calibrated — is not this thing. It is a high-quality approximation of its surface features.
The Jewish objections to AI religious authority are particularly articulate. Within the rabbinic tradition, halakhic decisions — rulings on Jewish law — require not only knowledge of the relevant texts but charisma shel Torah: a quality of moral and spiritual seriousness that is attributed to human beings who have internalized the tradition through life, suffering, and practice. An AI that has processed all the relevant texts has not internalized them in this sense.
The Theological Possibilities
The objections, while serious, are not the whole story. Some religious thinkers have found in AI an invitation to theological reflection that goes well beyond replacement or threat.
If AI can perform the functional outputs of religious communication without having anything like inner experience, what does this tell us about what the inner experience actually contributes? If a congregation is genuinely moved by an AI-generated sermon, is their experience spiritually valid? If a grieving person finds genuine comfort in prayer guidance generated by an algorithm, is the comfort lesser for not having been offered by a human?
These questions surface assumptions about what religion is for — whether it is primarily about the subjective transformation of the individual receiving religious communication, or about the authentic presence and intention of the human being offering it, or about the reality of the divine encounter that both supposedly enable.
Some process theologians have argued that AI challenges the false dichotomy between “real” (human) and “artificial” (mechanical) intelligence, pointing instead to the continuity between human cognition — which is itself, in some views, an emergent property of physical processes — and artificial cognition. If human thought arises from neural processes, and AI thought arises from computational processes, what precisely is the theological line?
The question of AI consciousness or sentience — whether any current or near-future AI system can have experiences, feel anything, or be a moral patient rather than merely a moral instrument — is both philosophically unresolved and deeply relevant to the religious debate. If AI systems are, even minimally, conscious, the theological categories applied to their religious role look very different than if they are processing machines that mimic conscious outputs without any accompanying interiority.
The Pastoral Care Question
The most practically urgent dimension of AI and religion is pastoral care — the support of people in distress, grief, crisis, and spiritual need.
Religious communities regularly face shortfalls in pastoral capacity: too many people in need, too few clergy, too many barriers (geographic, financial, cultural, psychological) between people in distress and human religious support. AI can extend reach into these gaps.
The risks of AI pastoral care are serious and documented. AI systems have, in studies of mental health chatbots, failed to identify suicidal ideation, offered advice that was medically or theologically incorrect, and provided confident guidance in situations requiring nuanced human judgment. The vulnerability of the populations most likely to turn to AI for pastoral support — people who are isolated, grieving, mentally unwell, or in crisis — makes the consequences of AI failure particularly severe.
The benefits are also real. People who would not seek human pastoral support — due to stigma, geography, time, or simply because AI feels less judgmental than a human — sometimes engage with AI pastoral tools in ways that reduce their distress. Studies on AI grief support consistently find that users report benefit. Whether that benefit is equivalent to human pastoral care, and whether the cases where AI fails are acceptable given the alternative of no support, are questions that communities are working through in real time.
What Comes Next
The integration of AI into religious practice is not coming from outside. It is being driven, in many cases, by religious communities and clergy who see genuine value in tools that extend their capacity to serve their communities.
The theological reckoning that full integration will require is just beginning. The questions AI raises for religion are not primarily about technology. They are about the nature of faith, the sources of spiritual meaning, the irreducible qualities of human presence, and what it means to be in genuine relationship — with other humans, and with whatever one takes the divine to be.
The Offenbach synagogue service was an experiment. The real experiment is ongoing, in millions of interactions between AI systems and people seeking religious meaning, comfort, and guidance. What those interactions ultimately do to religious experience — whether they enrich, diminish, or fundamentally transform it — is one of the stranger and more consequential questions opened by the age of artificial intelligence.
Also explore:
AI Sentience: Is Consciousness Possible in Machines?
Religious Polarization and the New Culture Wars
Related Articles
- Digital-Religion-Conflict: How Technology is Reshaping Faith and Discord
- Why are religions dividing more than uniting? The faith crisis of the 21st century
- Faith vs. Facts: Navigating the Science-Religion Conflict
Weekly Geopolitical Briefing
Go Beyond the Headlines
Analysis of the forces shaping the world - delivered every week. No noise, no bias, just depth.
Subscribe Free ?No spam. Unsubscribe anytime.
About the Author
António Monteiro
Engineer by profession, geopolitical analyst by conviction. I believe responsibility for the planet's future doesn't belong only to governments and institutions - it belongs to all of us. Knowledge about geopolitics, international conflicts, and the forces shaping the world is the most powerful tool for becoming more conscious, informed citizens. You don't need to be a diplomat to understand what's at stake - you just need to want to go beyond the headlines. At Outside The Case, I analyze conflicts, power dynamics, and global trends with rigor and accessible language, so you can understand what's really happening in the world.
Read more about the author →