AI is an ultra-capable black box, and both AI and the people who use it have predictable tendencies that can lead us astray.

Faithful Automation is a set of four questions to ask yourself when discerning whether, and how, to use AI within your vocation. It isn’t a list of rules, and it doesn’t assume AI is good or bad. It’s meant to be used one task at a time.

Our reliance on Christ’s faithfulness and wisdom

The “faithful” in Faithful Automation is first Christ’s faithfulness, not ours.

On our own, we lack the wisdom to faithfully apply AI in our vocations. We are nothing apart from Christ (John 15:5). Yet Christ, in his perfect faithfulness, works through us (2 Timothy 2:13). For those of us who teach, that includes forming wisdom in our students. Because students grow wise by walking with the wise (Proverbs 13:20) and become like their teachers (Luke 6:40), our own formation matters. So, when we lack wisdom, we ask God, who gives generously (James 1:5), through prayer and meditation on his Word.

Vocation matters here, too. Each of us has callings (e.g., teacher, student, parent, nurse, software developer, church member), and God works through those callings to serve our neighbors. Whether AI makes a task faster is worth knowing, but the more important question is whether our use of AI keeps our work pointed toward the people it’s meant to serve.

Thus, the four questions below are an attempt to support Christians as we strive for faithfulness to God and to our neighbors. They don’t replace prayer or Scripture. If you don’t share this faith, you’re welcome here, and you’ll likely find the questions useful anyway. But it would be dishonest to hide where they come from.

The four questions

Note that the questions scale with the stakes. A quick, low-stakes question to a chatbot might take a few seconds of thought. Handing AI work that affects your learning, your job, or other people deserves more. With practice, the questions tend to become habits.

1. Intent

What is my task, its purpose, and my intent in using AI to assist in completing it or to automate it entirely?

Every task has a purpose beyond getting it done. Sometimes the point is the product (e.g., a clear email, an accurate report, working code). Sometimes the point is the process (e.g., learning a skill, practicing a craft, working through a hard problem, showing someone you care). Often it’s both, and knowing which tells you what’s at stake when you hand part of the work to AI.

Note the difference between having AI assist with a task and having it automate the task entirely. Some tasks can safely be automated. Others can take some help but shouldn’t be handed off completely, and some shouldn’t involve AI at all.

Then be honest with yourself about why you’re reaching for AI. Some reasons are healthy (e.g., saving time on tedious work, getting unstuck, catching mistakes, exploring ideas). Others deserve a second look, like avoiding effort you actually need to invest or appearing more knowledgeable than you are. Intent also includes honesty with others. If the people affected by your work would be surprised or disappointed to learn how you used AI, pay attention to that.

• Who is this for, and what do they actually need?
• Is the value in the finished product, in the doing, or both?
• Am I trying to do this task better, or trying to avoid it?
• Would I be comfortable explaining how I used AI to the people this work affects?

2. Truth

How am I anchoring truth, knowing AI is neither transparent nor inherently trustworthy?

AI hallucinates, sounds confident when it’s wrong, and tends to agree with you, and you can’t see where its knowledge comes from. It also has access to more information than anyone could read, it makes our lives easier, and it’s often designed to convince us and keep us coming back. Put together, AI can quietly distort what we take to be true.

Anchoring truth means treating AI output as a draft you still need to check. This matters most when you’re outside your expertise, because that’s when errors and flattery are hardest to spot.

For Christians, truth is more than accurate facts. Christ calls himself the truth (John 14:6), and our work, including our teaching and learning, is meant to be anchored in him. That doesn’t make fact-checking less important. If anything, it raises the stakes of being careless with what’s true.

• Verify what matters. Check important facts, numbers, quotes, and citations against reliable sources (and actually open the sources).
• Be wary of easy agreement. Ask your question without revealing your opinion, or ask AI to make the strongest case against your view.
• Match your checking to the stakes. A brainstorm needs little verification. A medical, legal, financial, or academic decision needs a lot, often including a knowledgeable person.
• Know your limits. If you couldn’t recognize a wrong answer, don’t rely on AI alone.

3. Stewardship

How am I stewarding my own thinking, my time, and my gifts, recognizing the formative nature of this task?

The work we do is formative. Writing clarifies our thinking, struggling through a hard problem builds skill, and making decisions builds judgment. Teachers know this well: you don’t really know something until you teach it, because planning, teaching, and assessing are what form a teacher into a master of the material. When AI does that work for us, we may gain speed but lose the growth the work would have produced. Writing with AI is different than writing without it, and the same goes for brainstorming, planning, and problem solving.

Stewardship also covers your time and your gifts. Time saved is a real good, and handing mundane work to AI can free you for work that matters more. But discerning what’s mundane and what’s meaningful is harder than it sounds. Often the parts of your work you find most meaningful are where your gifts get used and grow, and they’re also where much of the joy in your work comes from. Those are the parts to be slowest to hand off.

• Think first. Sketch your own ideas, outline, or approach before asking AI.
• Use AI more like a coach than a ghostwriter. Ask for hints, questions, and feedback rather than finished answers, especially while you’re learning.
• Protect the skills that matter. Keep practicing the abilities your work, studies, and life depend on.
• Own the result. Make sure you can explain and defend anything you use without AI’s help, and take responsibility for it.
• Automate the mundane, not the meaningful. Then ask what you’re doing with the time you saved.

4. Connection

How is my use of AI displacing human connection that would otherwise occur when completing this task?

Many tasks are also chances to connect (e.g., asking a colleague for help, talking a problem through with a friend, getting feedback from a teacher, mentoring someone newer). Going to a peer, tutor, or teacher is usually slower than going to AI. But that efficiency gain comes at the cost of connection, and of the chance to grow in relationship. AI is always available and never judgmental, which makes it easy to skip the people. Those small interactions build trust and belonging that no tool can replace.

Our vocations are about serving our neighbors, so this question deserves real attention. Some tasks are largely solo work, and AI displaces little. Others are relational at their core, and AI can quietly put distance between us and the people we’re called to serve.

• Am I asking AI something I would otherwise ask a person? What would we both miss?
• Is AI helping me serve and connect with others, or putting distance between us?
• How will the people affected by this work perceive my use of AI?

Putting it into practice

You don’t need to memorize the questions. Return to them whenever you’re deciding whether and how to use AI for a task, and pray for wisdom when the answer isn’t clear. Over time, you’ll likely learn where AI helps you and where it costs you something.

For a worked example, see how the questions apply to creating course content with AI. And for how this framework fits into a larger picture of what it means to be AI literate, see the AI Literacy Framework.

Up nextThe AI Literacy Framework →