AI Mood Surveillance for Churches

A disturbing article was shared by CT’s Gleanings newsletter this week: a company in Brazil is offering facial recognition software for use in churches. It could be used to “take attendance”—but also, to discover gender and average age, and even to note moods (“happy, sad, distressed, or scared”), uncover statistical patterns (“likely reasons for being late”), and even trigger alerts for when a “pastoral visit might be helpful.”

I have long expected AI to intersect church life—but I was thinking about things like ChatGPT being utilized for sermon (or essay!) writing, or to help write donation requests (or even tell you when to send donation requests). I never expected someone to seriously offer a service that monitors who comes to church and what mood they are in when they do—much less for a church take them up on it.

I haven’t dug deeper than simply reading the article, so there may be nuance to this story. But the website, and what it offers, raises a number of alarm bells in my head.

First is the all-too-obvious elephant in the room: security. The website is minimalistic. It features logo saying they adhere to data protection, but nothing about the actual methods for doing so. Other services will provide great detail on how they keep such data safe (for example, here’s Slack); but this app offers nothing.

Additionally, there’s no indication of the people behind it. My cynical side wonders about the origin of the company and the software. Even doing a Whois lookup reveals nothing about the owners of the site. So—you’re providing quite a bit of information about your church and its members. Where does the data go? It says information on an event is only retained for 5 hours—but how does anyone know the company can be trusted?

Second, are users really equipped to think through the security issues? The app isn’t providing any details to educate its users on that. I hypothesize many (most?) churches in the USA would never touch this or any product like it—but if a church were using it, would people in the church know it? are rights being violated? (“Oh, we only use it to count the number of people in worship on Sunday morning. Why should we say we’re using it? We don’t talk about what accounting package we use to log donations.”) Do people using the software understand the security issues, or think about them? We are all very quick to self-surveil, with very little thought given to how long the data lives, who sees it, who has access to it, and so forth. I recognize I’m pretty open on social media, and I also know that is a built in limiter to some of the places I can go and people I can talk to—at least, publicly. Others might not be recognizing those issues. Would missionaries who visit a church (even in a Zoom room) be told? be aware? know to care? is there reason to care?

Third, and perhaps more importantly—I wonder about the perceived need for this kind of software. Is this just a shiny new toy? Is it for “really big churches” who feel like this can help keep people from being “lost in the crowd”? Is there a real demand for this? And if so, why?

At least one Brazilian church is using the tech, saying “we use facial recognition to provide extra assistance to members who are not coming to services.” That makes a certain amount of sense, I suppose—in a world where people are in their homes because they haven’t returned to in-person services, an AI system might catch signs that other people in a zoom room might not. Might. (I’m not saying it’s a good idea, I’m just saying it might be workable.)

But it’s one thing for a company to be monitoring the moods of their employees (and mood surveillance has been shown to be unreliable). It’s another for a church to be monitoring its parishioners, who are not employees. Is the church so far gone that we choose to rely on AI to give an alert about necessary pastoral care? Are we afraid to ask how someone is doing? Are we so disconnected? I am reminded of Robert Putnam’s book, “Bowling Alone,” on which Moore & others were reflecting this week. But surely the solution is not to outsource emotional status inquiries to a computer.

I could see how this sort of thing might be sold in the context of a megachurch. Megachurches are built around large crowds, but without the infrastructure for social networks within a church (small groups, etc), people can get ‘lost in the crowd.’ When they feel sufficiently lost, or they disagree with the church, they may migrate to a new church, taking their tithes with them. If mood surveillance can help identify people who could then receive a bit of extra attention, that might keep them from leaving—and the charitable gifts received could, in a sense, pay for the software. I’d bet that’s one of the arguments for the software, in any event.

But this seems to me just an extension of the idea that it takes a professional, trained, prepared clergy to serve a congregation. To multiply their ability to take care of a large flock, we give them tools to identify who needs extra attention—before they might leave the church because they didn’t get something taken care of. In order to preserve the congregation, we are going to institute “watching—monitoring—surveillance”? I shiver. The alternative is to trust interpersonal relationships of disciples to surface issues and respond to them.

Obviously, I’m not a Luddite. I use a tech tools in my work (here’s my list). I’ve explored the use of AI tools. I write my own articles, but I’ve explored ChatGPT a bit, and I use AI-based tools to generate some additional ideas around certain concepts and for things like grammar checking and email writing. I’d love an AI tool that can generate good headlines. But I would draw a very hard line at surveillance, and especially mood surveillance.

Regardless of your preferences, and mine, we should realize that surveillance—in some form or another—is coming for the mainstream church, and sooner rather than later. We in the mission field would do well to consider the implications of the camera, the livestream, and the footage archive. Those in churches with a foot in the mission world (i.e. mission pastors, mission boards, mission advocates) would do well to put a few cents in on conversations about such technologies.