The Seminary Intern Model:
Kenny Jahng's Framework
for AI in Church Staffing.

91% of church leaders support using AI in ministry. But support is not strategy. The gap between “we think AI is fine” and “here is exactly how we deploy it” is where the danger lives. Kenny Jahng filled that gap with one metaphor that changes everything.

AI's role in church staffing is that of “a seminary intern — beneficial as an assistant but necessitating faithful human oversight.”

That sentence, published through the National Association of Evangelicals in 2025, is the most practical deployment framework any church leader has produced. And once you understand it, you will never look at AI the same way.

What an Intern Does

Think about the best seminary intern you ever had. Smart. Eager. Good with research. Quick with a first draft. Reliable on logistics.

Now think about what you would never let that intern do unsupervised. You would never send them to counsel a suicidal congregant alone. You would never let them preach without reviewing the sermon. You would never hand them the keys to your church's doctrinal positions and say, “You decide.”

That instinct — the one that says “help, don't lead” — is the augmentation principle applied to AI. And most churches are violating it every week without realizing it.

The Two Columns

The intern model draws a clean line through every ministry function. On one side: tasks you can safely delegate. On the other: sacred ground where human presence is non-negotiable.

The delegable column is where AI saves churches hundreds of hours. The sacred ground column is where AI, left unsupervised, can cause irreversible spiritual harm.

Lifeway Research confirmed this with a survey finding that 95% of pastors affirm “discipleship is not completed through a program but is accomplished within relationships.” That is near-universal consensus. And it sets the theological floor for AI's role: AI can support relational ministry, but it cannot be relational ministry.

What Harvard Found About Blind Adoption

The intern model is not just pastoral wisdom. It is backed by the largest study on AI performance in professional settings.

Harvard Business School and BCG studied 758 consultants using GPT-4. Inside AI's capabilities, users completed 12.2% more tasks, 25.1% faster, at 40% higher quality. Impressive.

But outside AI's capabilities? Users were 19% less likely to produce correct solutions. And here is the critical finding: “Professionals who had negative performance when using AI tended to blindly adopt its output and interrogate it less.”

Apply that to ministry. AI inside its capabilities: research, scheduling, drafting. AI outside its capabilities: pastoral discernment, doctrinal evaluation, spiritual sensitivity. The professionals in the Harvard study who failed were the ones who treated AI like a senior partner instead of an intern. They stopped interrogating. They stopped supervising. They trusted the output and delivered it.

That is exactly what happens when a pastor copies a ChatGPT sermon outline without rewriting it from personal study.

Stoddart's Three Postures

Eric Stoddart's 2023 paper in Studies in Christian Ethics gives the intern model its theological spine. He identifies three postures a church can take toward AI:

The intern model maps perfectly: collaboration is the goal, delegation is acceptable with oversight, abdication is the line that must never be crossed. Every time you hand a ministry task to AI, ask: “Would I let a first-year seminary intern do this alone?”

If the answer is no, the AI doesn't do it alone either.

The Domain Expertise Requirement

Here is what makes ministry different from every other profession. In software development, GitHub Copilot made developers 55.8% faster — and only 1% relied exclusively on AI. Senior developers leveraged AI most effectively because they had the deepest domain expertise. They knew which suggestions to accept and which to reject.

The same principle holds in ministry, but with higher stakes. A developer who accepts a bad AI suggestion ships a bug. A pastor who accepts bad AI theology ships a heresy. The domain expertise required to supervise AI in ministry is not technical skill — it is theological training, pastoral wisdom, and Spirit-led discernment.

“Drawing on the gift of AI systems while humans retain their responsibilities for one another.”

Biola University's Good Book Blog put it in pneumatological terms: “AI can echo the structure of the gospel message, but it cannot comprehend it; only the Spirit of God reveals the things freely given to us by God.” This is not a claim about AI's current limitations. It is a claim about AI's permanent limitations. Understanding Scripture requires illumination. Illumination comes from the Spirit. The Spirit works through human vessels, not algorithms.

Deploying the Intern

For any church ready to move from “we support AI” to “here is our policy,” the intern model provides the deployment framework:

For every AI use case, ask three questions: (1) Would I let an intern do this unsupervised? (2) Does a human with theological training review the output before it reaches anyone? (3) Is this supporting relational ministry or replacing it?

If you answer “no, no, replacing” — the AI does not touch that function. Period.

This is not anti-technology. This is pro-stewardship. 91% of church leaders support AI. The intern model shows them how to deploy it without losing what makes the church the church.

Where OpenLumin Fits

OpenLumin was built to be the best intern a pastor could ask for. It does research. It surfaces evidence from 15+ scholarly sources — commentaries, historical context, original-language data, cross-references, Theographic Bible data covering 3,000+ people, places, and events. Every claim is sourced and marked as verified or flagged for review.

It does not preach. It does not counsel. It does not decide doctrine. It retrieves the evidence and puts it in your hands. You do the thinking. You do the praying. You do the preaching.

Because a good intern knows the difference between research and revelation.

About: AI Fluency Ministry helps the church understand and use AI wisely. OpenLumin is the practical application of that research — a free Bible research companion that retrieves evidence so you can do the thinking.