The Church Enters the AI Age

Leslie Taylor prepares sermons in Japanese for a bilingual congregation in Matsudo, outside Tokyo. Japanese is not simply a language she passes through on the way to ministry; it is part of the ministry itself. Years of study have given her enough competence to notice the difference between a sentence that is grammatically plausible and one that actually sounds right to the people who will hear it. When generative AI became widely available, Taylor found that it could help with difficult sentence structures, alternative phrasings, and nuances that might otherwise consume significant preparation time. She also found that it could be wrong in ways a less competent speaker might not recognize.¹

That small fact contains much of the argument of this book. Artificial intelligence can be genuinely useful before it is reliable enough to deserve independent authority. Its usefulness often grows when a capable human can evaluate it. And the very existence of a powerful tool does not settle who remains responsible for the result.

The Church has entered an age in which capabilities once restricted by money, geography, or specialized training are moving toward ordinary users. A missionary can rehearse a difficult conversation in another language at midnight. A small church can draft a policy, summarize a long report, generate accessible versions of teaching material, or build a simple software prototype without hiring a specialist for every step. A translation team can compare terminology across a corpus. A theological educator can prepare exercises for students with widely different levels of prior knowledge. A mission researcher can interrogate multiple datasets in ways that previously required significant technical assistance.

It would be easy to describe this only as a productivity revolution. That would be too small. The more important change is a redistribution of capability. The people who can attempt a task are no longer identical to the people who could afford or recruit the specialist labor required to attempt it before. In some settings, AI makes an existing workflow faster. In others, it moves a project from impossible to possible.

That possibility matters for Christian mission because missionary work often attends to needs commercial markets have little reason to prioritize. A low-resource language may represent a small addressable market. A church in a restricted context may need offline software designed around security assumptions very different from those of a consumer app. A theological school may serve students who cannot purchase large libraries. A small mission organization may need analysis, design, translation, or coding capacity that would normally exceed its budget. The fact that a technology can lower such barriers deserves more than suspicion.

At the same time, Christian mission cannot be reduced to the distribution of religious information. A person can receive accurate Christian content without becoming a disciple. A translation can be linguistically fluent and still be culturally inappropriate or theologically wrong. A chatbot can answer a pastoral question while remaining outside the structures of accountability that make pastoral authority meaningful. A ministry can automate administrative work and then discover that the time saved has merely produced a larger reporting burden. A powerful system can strengthen local Christians—or deepen their dependence on external platforms, languages, and institutions.

The central question is therefore not whether Christians should use artificial intelligence. They already do, and the category is too broad for a single yes or no. The better question is how Christians should steward forms of machine assistance whose capabilities are real, uneven, rapidly changing, and increasingly embedded in ordinary work.

This book approaches that question from within historic Christian conviction. God is not one variable in a technology strategy. Christ is not the patron of efficiency. The Holy Spirit is not a metaphor for human creativity. The Church is not merely a distribution network for religious information. Christian mission originates in God and is entrusted to a people who bear witness, make disciples, teach, baptize, form churches, serve neighbors, cross boundaries, suffer, persevere, and depend upon God.

These claims do not make technical questions irrelevant. They make them morally situated. If mission matters, then translation quality matters. If people bear the image of God, then privacy and dignity matter. If truthfulness belongs to Christian witness, then fabricated citations, synthetic testimony, and misleading media matter. If the Church bears responsibility for doctrine and discipleship, then systems that sound authoritative matter. If Christian mission is global, then the languages and communities systematically underserved by technology matter.

The result is a position more demanding than either enthusiasm or rejection. Christians should use AI more strategically where it removes genuine barriers. They should also become more discerning as capability increases. Greater capacity is not an argument for less supervision. In consequential work, it is often an argument for clearer supervision, stronger verification, more explicit authority boundaries, and greater attention to the people who bear the consequences when the system fails.

A recurring temptation in discussions of AI is to make the machine the protagonist. The story becomes a sequence of escalating technical abilities: language models, multimodal systems, agents, autonomy, perhaps eventually artificial general intelligence. Christian theology begins elsewhere. The important subject is not what the machine may become in the abstract, but what the Church is called to be and do in the presence of new tools.

That is why this book does not depend on resolving whether a future machine could be conscious, morally considerable, or intellectually superior to humans in some comprehensive sense. Those questions may become important. They are not required to answer the practical Christian questions already before us. A system does not need consciousness to expose sensitive data. It does not need a soul to produce persuasive theological error. It does not need moral agency to be granted dangerous permissions by a human organization. And it does not need to become a person before people begin treating it as a source of relational or spiritual authority.

The theological task is therefore both humbler and more urgent. We need to know what human responsibility cannot be surrendered merely because an output can be generated. We need to distinguish assistance from authority, production from judgment, information from wisdom, communication from witness, and efficiency from faithfulness. We need to recognize that some forms of friction are needless and should be removed, while others are bound up with learning, presence, patience, or formation and cannot be discarded without loss.

The missionary task also requires strategic imagination. If AI can lower the cost of language technology, theological education, software, accessibility, and research, Christians should ask where that capability can serve people who have historically received the least. There is no special virtue in preserving scarcity. The question is whether the technology can be developed and used in ways that strengthen local Christian agency rather than converting communities into permanent consumers of systems they cannot evaluate, govern, maintain, or refuse.

The governing conviction of this book can be stated simply: mission governs technology. Technology does not define the Church’s mission, decide what faithfulness means, or inherit spiritual responsibility. It remains an instrument within a larger Christian account of creation, fall, redemption, wisdom, community, vocation, and obedience.

That does not make the instrument unimportant. A tool capable of reshaping how people think, communicate, learn, build, and decide deserves serious attention. But the Church enters the AI age with a commission older than any machine and a responsibility no machine can carry on its behalf.

The tools have changed. The question is what we will do with them.

A Capability Shift

Earlier mission technologies were often visibly specialized. A printing press printed. A radio transmitted. Bible software searched a corpus. Generative AI is different because the same interface can cross task boundaries. One system may help with language practice in the morning, code in the afternoon, and theological research in the evening.

General-purpose capability is strategically valuable in small organizations. A missionary team rarely has enough specialists for every need. A system that offers mediocre-to-strong assistance across ten domains can sometimes matter more than a tool that is world-class in one domain the team rarely uses.

This is also why general-purpose systems deserve stronger judgment. Users can carry confidence from one successful task into another where the model is much weaker. A missionary who has received excellent writing assistance may assume the system’s historical research is equally reliable. A developer who has seen strong coding performance may overtrust advice about pastoral practice. The seamless interface hides uneven competence.

Christian AI literacy therefore needs transfer resistance: the habit of asking whether success in one domain actually justifies trust in the next.

The missionary opportunity is real precisely because the system is broad. Translation, research, accessibility, analysis, communication, and software can increasingly sit within one toolkit. But stewardship means refusing to let one friendly interface erase the boundaries among those responsibilities.

A common story of technological progress assumes that greater capability reduces responsibility. Better automation means less human involvement. For low-consequence mechanical tasks, that may be exactly right.

In ministry, however, greater capability can enlarge the domain for which a person or institution must exercise judgment. A small church that could never produce synthetic video had no reason to adopt a synthetic-media policy. Once it can produce realistic media in minutes, new responsibilities exist. A missionary who could not process thousands of records did not need to decide what patterns should govern automated targeting. Once analysis becomes accessible, the strategic and ethical choice becomes real.

Capability creates options. Options create stewardship.

This is a recurring biblical pattern even outside technology. Gifts, resources, authority, knowledge, and opportunity create corresponding responsibility. AI belongs inside that logic. The question is not whether the Church asked for every capability now arriving. The question is how faithfully it receives and limits them.

Where the New Capacity Matters Most

The institutional significance of generative AI appeared before most churches or mission agencies had vocabulary for it. Staff began using public chat systems to rewrite donor communication, translate messages, summarize reports, prepare teaching material, generate images, and troubleshoot software. Some leaders discovered the usage only after confidential information had already crossed into systems the organization had never evaluated.

This pattern matters because technological adoption is often bottom-up. Formal strategy arrives after convenience. A worker facing a three-hour task sees a tool that can finish the first draft in five minutes and reasonably tries it. By the time leadership debates whether AI belongs in ministry, the more practical question may already be which uses have become habitual.

That sequence changes the governance problem. A policy cannot be built only around future possibilities. It has to begin with ordinary work: what people are already doing, why they are doing it, what data they are using, and which parts of the workflow genuinely improve.

Christian organizations should therefore resist two instincts. The first is institutional denial—pretending AI use can be prevented simply by never mentioning it. The second is institutional enthusiasm—requiring adoption before workers have learned what the systems can and cannot reliably do.

The healthier approach is deliberate discovery. Identify real tasks. Compare old and new workflows. Ask what improved, what review burden appeared, what competence is required, and what risks changed. Use those observations to decide where standardization helps and where human judgment should remain primary.

The goal is not to make every Christian worker an AI enthusiast. It is to make hidden technological dependence visible enough to govern.

The strongest moral case for Christian AI investment appears where capability has historically been scarce.

A wealthy English-speaking church already has access to pastors, books, software, online courses, designers, and professional services. AI can still save time. The marginal benefit may be convenience.

A theological student studying in a language with few digital resources faces a different baseline. A church with no professional designer, no software developer, and limited access to translated material faces a different baseline. A Deaf community waiting for specialized translation tools faces a different baseline.

The same technical improvement can therefore have radically different human significance. This should influence Christian funding. Mission technology budgets often follow organizational visibility: polished websites, donor systems, headquarters analytics. These needs are legitimate. But AI also creates a chance to fund less visible infrastructure whose commercial return is small and missionary value is high: speech data for minority languages, accessible theological libraries, offline search, transcription tools, open linguistic resources, and secure systems for contexts commercial products do not prioritize.

The Church should be wary of an AI strategy whose largest benefit is making the already well-resourced slightly faster.

Stewardship Before Hype

Artificial intelligence is unusually vulnerable to both hype and backlash because demos compress complexity. A model generates a translation instantly, and the demonstration appears to prove translation has been solved. An agent completes a coding task, and the demonstration appears to prove software engineering has been automated.

Mission leaders should cultivate a simple discipline: distinguish demonstration from deployment. A demonstration asks whether something can happen once under favorable conditions. Deployment asks whether it happens reliably across ordinary cases, over time, under the constraints of real users, with acceptable error, cost, maintenance, and risk.

This distinction protects churches from chasing technology cycles and protects legitimate innovation from being discredited by overclaiming.

The book therefore treats spectacular capability as a reason to investigate, not a reason to declare victory.

A Mission Older Than the Machine

Footnotes

1. Ross et al., “Christians Use AI to Share Jesus.” ↩