The AI question churches are getting wrong

The loud fear is that AI will replace the pastor. The quieter and more common failure is asking software to do the part of ministry that only a present human can do.

Key takeaways

• The framing is the problem. The question most churches ask, "will AI replace pastoral work," sets up a yes-or-no fight that misses the two ways churches actually go wrong.
• There are two real failure modes, not one. A church can reject useful tools wholesale and bury its pastors in busywork, or it can expect software to carry the relational load only people can carry. Both leave people less cared for.
• The honest position is assistive. Software should take the busywork and surface who is drifting so a pastor has more time for presence. The tool prepares; the pastor decides.
• That is FlockConnect's posture. Collie, the built-in assistant, surfaces who looks isolated and drafts a note, but it never sends a message, writes to a record, or changes anyone's care on its own. A person reviews and approves every action.
• Some work is not delegable. Sitting with the grieving, knowing a name and a story, bearing a burden: no tool does these, and a church that hands them to software has confused efficiency with ministry.

The question is wrong before anyone answers it

Ask a room of pastors whether AI will replace pastoral work and the room splits in half. One side is alarmed, the other is curious, and the argument runs for an hour without anyone's church changing on Monday. The question feels urgent, but it is built wrong, because it offers two answers and both are bad: either AI takes over the work, or AI is dangerous and the church should stay away.

Real churches fail in a different way. They fail by rejecting tools that would have given a pastor back ten hours a week, so the pastor spends those hours on data entry instead of people. Or they fail by reaching for software to do the relational work itself, expecting an app to notice a hurting member, reach out, and somehow carry the relationship. The first church is exhausted. The second is automated and lonely. Neither problem is the one the headline question is about.

So the better question is narrow and practical. What is the work a pastor should never hand off, and what is the work a tool can carry so the pastor has more room for the first kind?

The two ways a church gets AI wrong

Rejecting the tools and drowning in the busywork

A pastor's week fills with work that has nothing to do with shepherding. Someone has to reconcile two member lists that disagree. Someone has to notice that a family has not been seen in a month, which means someone has to be holding the attendance pattern in their head. Someone has to draft the third follow-up note of the day. None of this is pastoral care. All of it eats the hours pastoral care needs.

A church that refuses every tool on principle does not protect its people. It just guarantees the pastor meets fewer of them, because the calendar is full of tasks a machine could have done in seconds. Suspicion of technology can feel like faithfulness. Often it is just a slower road to burnout, and the people who pay for it are the ones who quietly drift while the pastor is busy with a spreadsheet.

Expecting software to do the relational work

The opposite mistake is louder and more tempting. If a tool can flag who is isolated, why not let it reach out? If it can draft a warm message, why not let it send the message and close the loop? The logic is clean and the result is hollow. A person who learns that the "checking in on you" text was generated and sent without a human ever thinking about them has not been cared for. They have been processed.

This is the line that matters, and it is a line about what care is, not about capability. Care is one person choosing to know another and to act on that knowledge. A tool can surface the signal. It cannot supply the choosing, and the moment it pretends to, the relationship it was meant to support is gone.

The honest middle: tools assist, people decide

The position that holds up is unglamorous. A tool should do the work that does not require a soul, and it should hand everything else to a person at the right moment. It should take the busywork, watch the patterns no one can hold in their head, and put the right name in front of the right pastor while there is still time to act. Then it should stop and wait, because the next step belongs to a human.

That sounds like a small ambition. It is the right one. The goal is not a church that runs itself. The goal is a pastor with the margin to be present, because the hours that used to vanish into administration are now spent in living rooms and hospital hallways. Technology earns its place in a church by giving ministry back its time, not by doing ministry.

The companion idea is restraint about what these tools actually know. A flag that a member looks isolated is a prompt to go find out, not a verdict. The pastor who walks into the conversation may discover the member is fine, or is struggling for a reason no dataset could show. The tool was never the authority. It was the reason the pastor got there before it was too late.

How FlockConnect holds this line

FlockConnect was built around this exact posture, and the design says no to itself on purpose. It is a Church Relationship Manager, pastor-facing, that reads the signals a church already produces into a clear per-person view: who is connected, who looks isolated, and who has quietly slipped from the first group toward the second. For the working definition of the relationship it is trying to surface, what is a church connection is the place to start.

Collie, the built-in assistant, is the part people ask about, and the answer is deliberately modest. Collie can surface who looks isolated and draft a note or a suggested next step. Collie does not send the message, does not write to your records, and does not change anyone's care on its own. A person reviews and approves every action. The tool prepares; the pastor decides. That is the whole point, not a limitation the product is apologizing for. The assistant exists to get a pastor to the doorstep faster, never to knock for them.

The rest of the design follows the same instinct. FlockConnect works with what a church already has, offering an official two-way Planning Center integration as its one native connection and CSV import for everyone else, so a church does not have to leave a system it trusts to start seeing connection. It is priced by church size, not per seat, with a free trial, so the people who serve a church are never the line item. None of that is the heart of it, though. The heart of it is that the software is built to make a human more present, not less needed.

What software cannot do, and was never meant to

There is a category of ministry that no tool touches, and naming it keeps the rest in proportion. A pastor sits with a widow on the worst night of her life and says nothing useful, because there is nothing useful to say, and the sitting is the ministry. A pastor learns a teenager's name and the thing that teenager is afraid of, and remembers both six months later. A pastor carries a confession that cost someone everything to speak aloud. These are not inefficiencies waiting for a better workflow. They are the work.

C.S. Lewis pressed on why this is in his sermon "The Weight of Glory," where he argues that there are no ordinary people, that every person a Christian meets is a being of staggering and eternal weight, and that next to a community of such people the most impressive institution is a passing thing. If that is true, then a person is not a record to be managed or a signal to be processed. A person is the most serious thing in the room. Software can help a pastor find that person sooner. It cannot stand in the place where the pastor then has to be.

Francis Chan makes a kindred point about the shape of the church itself. In Letters to the Church he argues that the New Testament describes a family before it describes a program, and a family is not run by an algorithm assigning who belongs with whom. It is built by people who choose one another, slowly, in person. A tool can lower the odds that someone is overlooked in that family. The belonging itself is still handmade, and the church that forgets this trades something it cannot get back. For more on how that kind of community actually forms and why it shapes discipleship, see the early church model.

A simple test before adopting any tool

A church does not need a committee to sort the good uses from the bad. One question does most of the work. Does this tool free a person to be more present, or does it stand in for the person?

Run the candidates through it. A tool that reconciles two member lists frees a person. A tool that drafts a note for a pastor to read, edit, and decide whether to send frees a person. A tool that flags a quiet member so a real human can reach out frees a person. A tool that sends the message itself, or decides who gets care, or counsels a struggling member in place of a pastor, has crossed from assisting into replacing, and the church that adopts it will be more efficient and less of a church. The test is not whether the tool is capable. It usually is. The test is whether a human still does the part that only a human can do.

About the author

Michael Tribett is the founder of FlockConnect, a Church Relationship Manager built to help pastors see who is connected and who is drifting. He holds a Master of Divinity in Christian Ministry from Southeastern Baptist Theological Seminary, where he focused on missions and discipleship, and he serves as a small group leader at his church in the Raleigh, North Carolina area. FlockConnect is an official Planning Center partner.

Frequently asked questions

Will AI replace pastoral care? No, and the question itself is the wrong frame. The work pastors hand off well is the busywork: list reconciliation, pattern tracking, draft notes. The work that defines pastoral care, presence with the suffering and knowing people by name and story, is not a task a tool can carry. Refusing tools does not save that work; it just stops ministry from getting done.

What is the real mistake churches make with AI? There are two. Some reject useful tools wholesale and leave pastors buried in administrative work, which means fewer people get cared for. Others expect software to do the relational work itself, sending the messages and managing the care, which produces something that looks like attention but is not. The healthy path is between them: let tools take the busywork, let people do the relationship.

Does FlockConnect's assistant message members on its own? No. Collie surfaces who looks isolated and can draft a note or a suggested next step, but it never sends a message, writes to a record, or changes anyone's care on its own. A person reviews and approves every action. The tool prepares; the pastor decides.

How should a church decide whether an AI tool is appropriate? Ask one question of any tool: does it free a person to be more present, or does it stand in for the person? Tools that take busywork and surface the right name at the right moment free a person. Tools that send the care or make the relational decisions replace the person, and that is the line to hold.

Is it unfaithful for a church to use technology at all? No. Refusing every tool does not protect a congregation; it usually just guarantees the pastor spends ministry hours on tasks a machine could do in seconds. Used well, technology gives time back to the human work. The danger is letting tools do the part that was always meant to be human.

What does FlockConnect actually do for a pastor? It reads the signals a church already produces into a per-person view of who is connected and who is drifting, works alongside the system a church already runs through an official Planning Center integration or CSV import, and uses Collie to surface and draft, never to act on its own. The point is to put the right person in front of a pastor at the right time so a real relationship can do the work software cannot.