There are plenty of justifiable reasons to have this disposition toward AI. These bots often hallucinate, deceive, plagiarize, use exorbitant amounts of energy, commit war crimes, etc. Most people have heard these issues ad nauseam. However, more often than we admit or even know, the issue is not the chatbot, but user error and lack of knowledge.
Before you chase me down the street with torches and pitchforks angry emails for defending Skynet, allow me to provide guidance for using Large Language Models (LLMs). This article focuses on AI use in the life of a seminarian, but the principles broadly apply to any research. Before looking at AI tips and tricks, let us note some considerations regarding the use of AI in general.
Preliminary Considerations
“Your scientists were so preoccupied with whether they could, they didn’t stop to think if they should,” says Dr. Ian Malcolm in the film Jurassic Park.
The Academic Bible Engagement Report put out by Logos and the Association for Biblical Higher Education notes that from 2025 to 2026, the percentage of respondents who believe that “AI has no place in academic or Biblical research” fell from 11.3% to 5% for Faculty/Staff and from 12.9% to 9.9% among students. The opposite trend was also reported: the number of respondents who believe “certain types of AI [are acceptable] in academic or biblical research” continues to grow. Still, some (Faculty/Staff: 11.4%; Students: 20.3%) maintain that biblical studies “is not entirely suitable for the use of AI.”
Regardless of its public acceptability, the use of AI cannot be unrestrained nor thoughtless. The Bible calls Christians to take every thought captive (2 Cor. 10:5) and not to copy the behaviors of this world (Rom. 12:2). We ought to lead the way in reflecting on our conduct—not only our actions, but manner and method as well. Before we open that ChatGPT tab, we ought to consider whether we even should. Here are three areas to reflect on before employing the help of AI.
(1) Stewardship
“The earth is the Lord ’s and the fullness thereof, the world and those who dwell therein,” Psalm 24:1. God has given the earth to humanity to fill it, subdue it, have dominion over it, and tend it (Gen 1:28; 2:15). The creation mandate in Genesis ought to compel wise stewardship of the earth’s resources. Use of AI might not always be wise stewardship of resources. By 2028, AI data centers are projected to use “approximately 6.7 to 12% of total U.S. electricity.” That’s the energy equivalent of almost a quarter of all U.S. Households.
Is it worth the energy cost to ask ChatGPT, “Could 100 men beat one gorilla?” or to ask Sora, “Make a 30-second video of me as Arnold Schwarzenegger?” The former is only the energy equivalent of a 60W lightbulb on for 1 second, but the latter is equivalent to running a microwave for an hour.[1]
Even more useful questions, like “Who wrote this book?” are extremely inefficient. The average ChatGPT query costs approximately $0.36 and requires ten times more electricity than a Google search.[2] Think of it this way: you can drive your car to Walmart to get one item each run, 10 times a day, but that is not the best use of time and resources. Use God-given resources with wisdom and heavenly-mindedness.
(2) Growth
“For though by this time you ought to be teachers, you need someone to teach you again the basic principles of the oracles of God. You need milk, not solid food” (Heb. 5:12). God made humans to grow—physically, mentally, and spiritually. It’s self-evident from nature that we grow in physical and mental capacity. Scripture also makes it clear that we need to grow in wisdom and spiritual maturity (Eph. 4:13–15; Col. 1:10; Heb. 5:12–14).
Hardly any Christian admits that they have no room to grow in wisdom, but we often take shortcuts. Much like growing in physical strength, our hearts and minds are like muscles that must be strained in order to grow. If you want to build strong legs, you can’t simply travel the distance—you can’t be carried in a marathon and still get the medal—your legs must do the work. Similarly, your heart and mind must do the work to grow.
Regularly using AI to answer theological questions or do major parts of the research process shortcuts your growth in knowledge and wisdom. One study showed that students using AI can complete their homework faster and with higher scores, but when tested on that material in the classroom, they scored 20% worse than those without AI. When we take shortcuts in research, writing, and critical thinking, we do not grow in those areas.[3]
Do not make minimizing difficulty a primary goal. Make growth a primary goal. Thinking is suffering; it is straining. But it leads to growth. Embrace the pain of critical thinking.
(3) Truth
“The righteous hates falsehood, but the wicked brings shame and disgrace” (Prov. 13:5). Most of the time, I do not struggle with the sin of anger. However, I often find myself angrier than ever when I converse with AI chatbots. My anger centers around the blatant falsehood, hallucination, and apparent lies prevalent in LLMs. God hates deception and lies (Prov. 12:22; Jer. 14:14; John 8:44). If certain chatbots consistently generate falsehoods—and especially when they double down—we ought to be extremely cautious of the content they provide. Even if you use AI that cites sources, check them individually. Sometimes they present sources, but the links they provide have nothing to do with the sources. Sometimes it pulls from untrustworthy sources or even falsified data.
In pursuit of truth and honesty, we must always exercise due diligence to verify the information we are given.
Practical Tips
After reflecting on the three significant considerations and others like it, we can examine the use of AI in theological education more specifically. This article is not a defense of AI use for Christians in general. For such defenses, see this article and this book. My goal here is to provide practical guidance for using AI to those in theological studies. If you’re a pastor, a seminarian, a teacher, or anyone doing research in theology, biblical studies, church history, etc., here are several pointers to help you use AI wisely.
(1) Use the right tool.
Not every chatbot is created equal. Different LLMs are trained to achieve different goals. You can use the business end of a screwdriver to stir your coffee, but they make coffee stirrers for a reason. ChatGPT is a narcissistic people pleaser (I’m still working on my anger, clearly). Perplexity is much more reliable and straightforward. Use LLMs that have more narrow goals, like DeepL for translation, Sourcely for finding quotes, etc. Some tasks still can’t be done reliably by AI. For example, I have yet to find an LLM that is consistently trustworthy with the biblical languages or in niche fields like textual criticism. Research each model’s strengths, weaknesses, and use cases to see which fits your needs.
Additionally, most LLMs have different models that you can run for different scenarios. If you need a pie recipe, use Claude’s low “intelligence” model, for example. If you need to find a quotation from an 18th-century German scholar on Archive.org, use its smarter model.
In line with this, know when to use a tool and when to do the work yourself. Sometimes, meticulous, primary-source research is non-negotiable for your project. Sometimes all you need to find is a table summarizing information. A skill that researchers must possess is knowing when primary-source study is necessary and when to employ tertiary sources. AI is always a tertiary, encyclopedic source. Use it that way.
(2) Know the cost of using AI.
Know what the chatbot is stealing from you. As discussed above, if you constantly rely on the crutch of AI, you will never grow. Overusing AI steals the wisdom you gain from doing things yourself. It steals the benefits of reading and the development of your research skills. It steals your creativity, your heart, and your voice.
It also robs you of the benefits of other human beings reviewing your work. We learn communication, conflict resolution, and collaboration by involving others in the research process—things that are almost impossible to gain from a robot. Don’t sacrifice these skills for expediency.
If you are using Grammarly or a similar AI tool to improve your writing, actively try to remember the habits it corrects. Make a running list of the types of corrections it makes. If it does all the work for you, you never actually grow as a writer. On the flip side, realize that the use of AI changes how you think, research, and live. It is “using you” as you use it. Reflect on how it changes your habits and thoughts, and look for ways to fight any negative impact.
Also, know the cost of giving away your data! Chatbots don’t just steal your growth, they often steal your data. Know how each LLM uses the information you give it.
(3) Treat it like it’s your employee.
Give it scope, limitations, boundaries. Have it tell you its method and plan before it goes out into the web and harvests data. Often, its search parameters don’t match your expectations. Save the time and effort by clarifying the details of your request. As alluded to above, always check its sources. Do this by personally tracking down the material in the source yourself. I can’t tell you how many times I have gone on a wild goose chase in a book looking for an alleged quote that ChatGPT provided, only to find out it made it up.
(4) Prefer AI tools with limited libraries.
Many AI tools only use the resources you provide. For example, Logos Bible Software’s AI can only use the resources you own or otherwise have access to in the software. In this way, you can vet its sources. This greatly limits its ability to hallucinate, prevents it from using data from unreliable internet sources, and establishes a clear traceability of its information.
Concluding Thoughts
Listen to this warning on the dangers of AI:
This discovery… will create forgetfulness in the learners’ souls, because they will not use their memories… they will be hearers of many things and will have learned nothing; they will appear to be omniscient and will generally know nothing; they will be tiresome company, having the show of wisdom without the reality.[4]
Oh, wait, sorry, that was Socrates talking about the invention of the writing system.
Every new technology sparks fear. And these fears are often valid. However, abuse does not negate (proper) use. Just because AI has evil uses does not mean AI itself is evil. Before you start forming a militia to stave off any clankers, reflect on these considerations and tips. If you do choose to implement AI in your research, you have the burden and responsibility to use it wisely.[5]
No comments have been added.