AivexaNewsSearch
AI news for builders and product teamsChecked every hour

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

Collected Oct 30, 2026

David Durman, an administrator at the University of Kentucky, has obtained a religious accommodation allowing him to abstain from using AI at work, according to an announcement the American Humanist Association published earlier this week. Durman is described as the first Humanist — and specifically a nontheistic one — to secure such an accommodation over AI use, and as one of the first people to publicly discuss having sought one. He told TechCrunch he was prompted to act after reading that Erin Maus, a Unitarian Universalist software engineer, had secured similar protections.

Durman's objection is not theological in the conventional sense. He describes Humanism as a set of beliefs about human experience and how people engage with the world, centered on well-being and flourishing for individuals and for the communities around them. His concern is that leaning on AI erodes the active use of one's own mind, which he sees as the mechanism by which people become moral actors toward themselves and others. In a statement carried by the AHA, he also pointed to AI's effect on people's ability to earn a living, its environmental costs, and what he characterized as its supplanting and theft of human creativity in areas such as art and literature.

That framing matters for how the accommodation was framed legally. In the United States, workplace religious accommodation generally rests on the requirement that employers reasonably accommodate an employee's sincerely held religious beliefs unless doing so imposes undue hardship on the business. Humanism is a nontheistic worldview, and the AHA's position is that nontheists fall within that protection alongside members of established faiths. Alexandria Huiskens, a staff attorney at the AHA, told TechCrunch that concern about AI and human dignity runs across traditions — Catholic teaching, Universalism, Humanism — and that the organization wants Humanists and other nontheists to know they are covered by workplace law too.

The university's response was narrow and case-specific. Spokesperson Jay Blanton said the University of Kentucky evaluates religious accommodation requests individually, weighing the employee's job responsibilities and the circumstances, and concluded in this instance that an accommodation could be granted while Durman continued to perform the essential functions of his role. Blanton added that accommodations may be revisited as technology and job duties evolve, so the arrangement is best understood as a point-in-time determination rather than a permanent carve-out.

Durman's role helps explain why the request was granted. He supports researchers applying for grants, an administrative position in which AI use is not currently mandatory; colleagues and leadership have suggested employees try the tools, and he sought the exemption preemptively in case a suggestion becomes a requirement. Maus's case is more striking precisely because software engineering is one of the fields where writing code without AI assistance has become unusual. The contrast suggests, as my own reading, that the practical reach of these accommodations will depend heavily on the job: roles with discretionary tool use are easy to carve out, while roles where AI is embedded in the production pipeline invite harder questions about what counts as an undue hardship.

This also illustrates a broader shift in how AI policy lands in organizations. Many employers are moving from encouragement to expectation — recommending assistants, then measuring adoption, then wiring tools into workflows. Each of those steps raises the stakes for employees with conscientious objections, and the mechanism they have available is the same one used for dietary, scheduling and dress requests: a sincerely held belief, disclosed to the employer, evaluated case by case. This suggests HR and legal teams should expect AI to join the list of topics that generate accommodation requests, and that documentation of job duties will matter when deciding whether abstention is workable.

The public nature of the case is part of its significance. Durman said his aim was personal — he is not telling others how to think — but he hopes it encourages people with similar concerns to pursue resolution wherever they judge it most appropriate. The AHA's involvement signals an intent to normalize the idea that secular worldviews qualify for the same protections as religious ones, which could widen the pool of employees who feel entitled to ask. Durman is not the first to raise the issue, but the pairing of a Humanist claimant with an institutional statement and a university's on-record confirmation gives the pattern a template.

Why it matters: Employers rolling out AI tools should assume some staff will frame reluctance as conscience rather than preference, and that some will seek formal accommodation. For developers and product teams, the immediate effect is likely procedural — clearer policies on when AI use is required versus recommended, and records of which tasks genuinely depend on it, since those records determine whether an exemption is reasonable. My inference is that this will mostly affect administrative and knowledge roles first, while engineering teams face harder trade-offs where AI is built into delivery.

Read at TechCrunch · AI

Based on reporting from the original publisher. Visit the source for full context and later updates.

Publisher excerpt

David Durman, a University of Kentucky administrator, recently earned a religious exemption from using AI in the workplace.