You don’t need to believe in a god to get a religious exemption from using AI at work

I have spent most of my career building things, usually at the intersection of where code meets commerce. Normally, when a new tool comes along, the debate is about efficiency, cost, or whether it actually works. But we just hit a new milestone in the friction between humans and machines, and it has nothing to do with technical specs. It is about the soul, or at least, the legal definition of one.

David Durman, a staffer at the University of Kentucky, recently successfully argued for a religious exemption to avoid using AI tools at work. Here is the kicker: he did not need to cite a specific deity or a centuries-old scripture to do it. He relied on a deeply held personal belief system that views AI as something fundamentally incompatible with his humanity. For anyone building AI startups or managing technical teams, this is a massive signal that the next big hurdle for adoption is not going to be the hardware; it is going to be the HR department.

The Non-Religious Religious Argument

In the United States, religious exemptions under Title VII are surprisingly broad. You do not have to belong to a mainstream church. You do not even have to believe in a god. The legal standard generally looks for beliefs that are "sincerely held" and address fundamental questions of life, purpose, and death. Durman effectively argued that delegating human thought and creative labor to a machine violates his core moral compass.

From a founder’s perspective, this is a nightmare and a fascinating case study all at once. We are used to dealing with people who are slow to adopt new tech because they are afraid of breaking things or losing their jobs. We are not used to people refusing to use a tool because they believe it is an existential threat to their moral integrity. This moves the conversation from "how do we train users" to "can we even require users to exist in this ecosystem?"

Why Builders Should Care

If you are building an AI-first company, your market just got a lot more complicated. For years, the assumption has been that once the software is good enough, everyone will use it because the efficiency gains are undeniable. But human beings are not always rational actors in the economic sense. We are emotional, spiritual, and often stubborn.

If employees can start opting out of AI workflows based on philosophical grounds, the entire promise of "AI-augmented labor" starts to crumble at the edges. Imagine you sell a B2B SaaS platform that automates 40% of a customer service rep's job. If 10% of that workforce gets an exemption, your ROI for that client drops. The client now has to maintain two parallel workflows: one for the machines and one for the "conscientious objectors." That is a massive operational tax.

The Authenticity Tax

Durman’s argument hinges on the idea that human interaction and human effort have an intrinsic value that cannot be replicated. As a builder, I actually find myself agreeing with him more than I thought I would. We have seen what happens when everything becomes automated—content becomes bland, interactions become robotic, and the "vibe" of a product disappears. The skepticism I hold about the current AI hype is that we are prioritizing speed over substance.

However, the legal precedent here is the real story. If a university administrator can opt out, who else can? Can a software engineer refuse to use Copilot? Can a doctor refuse to use an AI diagnostic tool because they believe healing requires a purely human connection? The doors are officially open for a new kind of Luddite movement, one protected by the same laws that protect Sunday worship.

The Management Dilemma

For those of you running teams, you need to look at your internal AI policies right now. Most companies are rushing to implement "AI policies" that focus on data privacy and security. Almost nobody is thinking about the "human rights" aspect of being forced to use generative models. If you force a tool on a team that views it as a moral affront, you are not just dealing with a productivity dip; you are looking at a potential lawsuit that you might actually lose.

The solution is not to double down on the tech, but to be honest about where the tech fits. We have to stop selling AI as a total replacement for human thought. When we frame it that way, we invite this kind of pushback. If we frame it as a utility—like a calculator or a spreadsheet—the religious argument becomes harder to sustain. But as long as we keep using words like "intelligence" and "consciousness" in our marketing, we are handing the skeptics the very ammunition they need to claim it is a spiritual issue.

What This Means for the Future

We are entering an era of "Cognitive Liberty." This is the idea that you have the right to control your own mental processes. If an AI tool is seen as an intrusion into that process, people will fight it. Durman’s victory at the University of Kentucky is a shot across the bow for the entire industry. It proves that the human element is not just a variable we can optimize away; it is a legal and ethical wall.

For founders, the takeaway is clear: stop building tools that try to be people. Build tools that help people be better. If the software feels like a soulless replacement, expect more people to find their "religion" real quick. The most successful AI companies of the next decade won't be the ones with the best LLMs; they will be the ones that figure out how to coexist with a workforce that is increasingly skeptical of the machine.

The Bottom Line

• Legal precedent is shifting: You don't need a church to claim a moral objection to AI.
• Operational drag: Opt-outs create fragmented workflows, killing the efficiency gains of automation.
• Marketing matters: Stop calling it "intelligence" if you don't want people to treat it like a spiritual rival.

We are building the future, but we have to remember that the people living in that future still want to feel like they matter. If we ignore that, no amount of VC funding or GPU clusters will save us from the human desire to remain human.

Read the original at TechCrunch AI →