AI Isn't the Threat. Intellectual Complacency Is.

Ask a room full of pastors what they think about artificial intelligence, and you will get two answers in the same breath. Most will tell you they are wary of it, even threatened by it. Many of those same pastors will also tell you they believe God can work through it anyway.⁸That is an old and recognizable posture, one the church has taken before, toward tools that reshape how people encounter the Bible.

This article is not an argument that AI is safe, or that concerns about it are overblown. Some of those concerns are serious and deserve to be taken as seriously as the people raising them intend. It is an argument that the question in front of the church right now, whether AI is compatible with authentic Christian faith, is not new in kind, even if it is new in scale and speed. Christians have faced a version of this question every time a technology changed who could read Scripture, and how. The pattern is worth looking at closely, because it complicates the easy story on both sides: neither “every new tool is a gift” nor “every new tool is a threat” survives contact with the historical record.

A pattern older than the internet

When Johannes Gutenberg printed his first Bible around 1455, he produced somewhere between 160 and 185 copies.¹ That number sounds modest until you know what it replaced. A single hand-copied Bible could take a scribe the better part of a year and cost more than most households would spend on a home. Within about fifty years of Gutenberg's press, Europe had produced more books than had existed on the entire continent in the previous thousand years combined.² Historian Elizabeth Eisenstein, whose 1979 study The Printing Press as an Agent of Change remains the standard scholarly account, put a typical edition from this era at 200 to 1,000 copies,³ reproduced identically, at a fraction of the cost, again and again.

The Bible sat close to the center of this shift, not off to its side. William Tyndale's English New Testament, translated directly from the Hebrew and Greek in the 1520s, took full advantage of print's ability to reproduce a text faster than any authority could suppress it.⁴ Producing an English Bible was illegal in England at the time. Tyndale worked from exile, was eventually betrayed, tried for heresy, and executed in 1536,⁴ strangled, then burned, for making Scripture legible to people who had never learned Latin.

Martin Luther's German New Testament, published a few years earlier in 1522, sold out its first run of three thousand copies within a week, despite costing more than a week's wages for an ordinary laborer. Twelve authorized reprints and sixty-six unauthorized ones followed within the year. By the time Luther died in 1546, roughly two hundred thousand copies were in circulation.⁵

By the nineteenth century, expanding access to Scripture had become an organized mission rather than a side effect of new technology. The British and Foreign Bible Society, founded in 1804, and the American Bible Society, founded in 1816, both stated their purpose almost identically: to circulate the Bible without note or comment, to as many people as possible, explicitly including those too poor to buy one on their own.⁶ Alexander Cruden's Complete Concordance, compiled single-handedly in 1737, let an ordinary reader trace a word or theme across the entire Bible without years of memorization or a personal library of commentaries; it has stayed in print ever since.¹¹ And today, YouVersion's Bible App has passed one billion device installs, with tens of millions of people opening it on an average day, the large majority outside the United States, in places where a printed Bible in one's own language is still not something to take for granted.⁷

None of this is a story about a single invention. It is a pattern going back to the printing press: nearly every major shift in how information moves has, sooner or later, been turned toward the specific project of putting Scripture into more hands, in more languages, at lower cost.

What the pattern actually shows

It would be convenient to end the story there: technology keeps expanding access, access is good, so more access is always better. But that reading leaves out the part of the history that makes it useful.

Every one of these shifts was also a fight. Tyndale is not a footnote to the printing press's triumph; he was killed by the authorities who understood exactly what an English-reading laity would mean for who got to interpret Scripture on their behalf. The Reformation's translation controversies were not a brief headache before everyone agreed printing was good. They were serious, sustained arguments about whether wider access would produce more faithful readers, or more confused and divided ones. Both things turned out to be true, depending on where and how you looked.

The version of this argument that says the church always resisted new technology and was always wrong to is its own kind of simplification, one that treats every objection as reactionary fear rather than considering that some of the concerns were substantive, and some of what followed vindicated them.

Wider access to Scripture did not, on its own, produce deeper discipleship or more careful reading. It produced remarkable movements of biblical literacy and centuries of church conflict over interpretation, sometimes from the very same technology. The tool did not settle the argument about how it should be used. It made the argument available to more people, faster.

Taking “this time is different” seriously

So does AI belong in this pattern, or does it break it?

The honest answer is that there is a real case for both, and the case against treating AI as just the next item on this list deserves to be stated plainly, not waved off.

Every technology in the pattern above distributed a fixed text. The words did not change; only the speed, cost, and reach of getting them to a reader changed. A printed Bible in 1522 said the same thing a hand-copied Bible said in 1400. An AI system answering a biblical question is doing something categorically different: producing new language, in real time, that did not previously exist, shaped by a process most users cannot inspect and few can fully explain. That is closer to an authorship problem than a distribution one, and authorship is precisely the terrain the church has always guarded most carefully, not because it fears expertise being shared, but because it has hard-won reasons to care who is speaking, and on what authority.

This is a legitimate distinction, and pastors who raise it deserve to have it taken on its own terms, apart from Tyndale's history. A concordance cannot misquote a verse it did not generate. A printed Bible cannot fabricate a citation. An AI system, depending on how it is built, sometimes can. Barna's 2024 survey found that seventy-two percent of pastors describe AI as a threat to the church, compared with fifty-seven percent of practicing Christians generally.⁸ The gap itself is informative. Pastors disproportionately are the people who have to answer for what gets said from the pulpit, in a Bible study, or in pastoral counsel. Their caution is often a professional judgment about accountability, applied to a genuinely new kind of tool, not simple fear of the unfamiliar.

At the same time, that same research found seventy-seven percent of pastors agreeing that God can work through AI anyway.¹² Separate Barna research found that AI-assisted sermon preparation among pastors nearly doubled in a year, with only thirteen percent of U.S. pastors reporting no AI use at all.⁹ Whatever the theological verdict on AI turns out to be, church leaders are not waiting for a settled answer before engaging with it. Real caution and real use, held together without much apparent contradiction, is closer to how the church has actually related to new technology throughout its history than either the alarmed or the triumphant version of the story suggests.

So where does that leave the authorship objection itself? Here is the actual position this article takes, not a hedge: the objection is correct about AI in general, and wrong about what AI is required to be. An AI system that generates a confident, unattributed answer to a biblical question really is doing something the printing press never did, and the church's caution about that specific design choice is earned, not reactionary. But an AI system built to cite what it draws on, to show which claim comes from the text itself, which comes from a particular scholar or tradition, and which is genuinely contested, is not authoring Scripture's meaning. It is doing something closer to what a concordance or a study Bible's footnotes have always done: surfacing material a reader could not easily gather alone, while leaving the reader able to check the sourcing rather than simply trust the voice. The authorship problem is real. It is also a design problem, not an inherent property of the technology, and that distinction is where the argument in this article actually lands: judge each tool by whether it is built to be checked, not by whether it is built on a language model.

The actual threat is complacency, not the tool

Here is where the historical pattern earns its keep. In every case above, the tool itself was never the whole story. What mattered just as much was what people did with the access once they had it.

A printed Bible in someone's hands did not guarantee they read it carefully, understood its context, or let it change how they lived. It guaranteed only that they could. The same is true of the concordance, and it is true of the Bible app: instant access to every occurrence of a word in Scripture does not, by itself, make someone a more careful reader. It makes careful reading possible in a way it was not possible before, for people who did not previously have the resources to attempt it. What they do with that possibility is a separate question, one the tool cannot answer for them.

AI raises the same question with higher stakes, because the tool is more capable of doing the thinking for you, if you let it. This is where the genuine risk actually lives: not in the existence of a tool that can generate a plausible-sounding answer to a biblical question, but in a posture, available in any generation, with any tool, that treats a plausible-sounding answer as a substitute for the work of sitting with the text, wrestling with its difficulty, and bringing it into conversation with a community of people who will push back. That posture predates AI by centuries. It shows up in memorization recited without being understood. It shows up in a congregation that trusts a pastor's word without ever opening a Bible to check it. It will show up, in some form, with whatever technology comes after AI. Calling it the AI problem gives the tool too much credit, and lets the older habit of mind off the hook.

The Stanford AI Index's most recent report found that generative AI reached an estimated fifty-three percent of the population within three years of its public release, faster than the personal computer or the internet took to reach the same milestone.¹⁰ Whatever else that speed means, it means the church does not have decades to develop wise habits of use before the tool is everywhere. It has, realistically, a much shorter window, closer to the one Tyndale's contemporaries had, measured in years rather than generations, to work out what faithful engagement with a text-transforming technology actually looks like.

Conclusion

None of this settles whether any particular AI tool, used in any particular way, is wise. That question deserves its own careful, case-by-case scrutiny, and healthy skepticism toward specific claims and specific products is warranted, arguably overdue, given how often new technology arrives promising more than it delivers.

But the broader question, whether AI's arrival is compatible with serious Christian engagement with Scripture at all, has, if the history here is any guide, a less dramatic answer than either side of the current debate wants to give. The church has navigated exactly this kind of disruption before, imperfectly, with real costs and real gains on both sides of the ledger. What made the difference each time was whether the people using the technology kept doing the harder, slower work it could never replace: reading closely, thinking carefully, staying in community with people who would say when they had gotten it wrong.

That work has never been automatic. The tool that makes it easiest to skip carries the same old danger every tool before it did, just in a newer shape, and it will take the same old discipline to use well.