A woman writes notes at her desk during a video call, with AI icons overlaid on the scene.
Every IT department in corporate America has spent the last few years wrestling with how to roll out generative artificial intelligence without breaking compliance, budgets, or the company network. Now, enterprise administrators and human resources teams have a fascinating new variable to factor into their deployment roadmaps: the workplace religious accommodation.

At the University of Kentucky, 45-year-old research administrator David Durman recently secured a formal Title VII religious exemption excusing him from personally using artificial intelligence tools—specifically naming Microsoft Teams Copilot, OpenAI's ChatGPT, and Anthropic's Claude—to complete his job duties. What makes the case particularly striking for employment law observers and IT leaders alike is that Durman's objection does not stem from a traditional theistic faith. Instead, Durman is a secular Humanist who successfully argued that outsourcing his cognitive and ethical labor to large language models violates his deeply held moral convictions.

While splashy headlines across the web have framed this as a sweeping "first-of-its-kind" legal victory against mandatory AI, the actual administrative details tell a much more nuanced—and far more practical—story for anyone managing Microsoft 365 environments and workplace software policies.

A woman writes notes at her desk during a video call, with AI icons overlaid on the scene. Inside the University of Kentucky Accommodation​

To understand what actually happened in Lexington, it helps to separate the viral framing from the administrative record. According to Religion News Service and the American Humanist Association (AHA), which provided legal support to Durman, the accommodation was not the result of a courtroom battle or a federal civil rights ruling. Rather, it was an internal workplace accommodation granted by the University of Kentucky’s Office of Equal Opportunity (OEO).

Furthermore, as Fortune and Dexerto noted, the University of Kentucky did not actually have a formal mandate requiring employees to use AI. Instead, Durman grew concerned by what he described to Fortune as the "growing normalization" of AI tools in his workplace, where leadership and colleagues offered mounting suggestions to integrate bots like Microsoft Teams Copilot into daily workflows and floated ideas about attending AI conventions. Hoping to preempt any transition from casual encouragement to a rigid job requirement, Durman submitted a formal accommodation request in late August 2026 with assistance from AHA staff attorney Alexandria Huiskens.

In his letter to the university, Durman cited his personal convictions alongside the 2003 Humanist Manifesto III, which emphasizes human dignity, informed choice, freedom of conscience, and personal responsibility.

"The process and work involved in thinking through situations, through my work, through my ethics, through the way I handle other people, is a key component in being able to develop myself as a moral person," Durman told Religion News Service. "And offloading that labor, regardless of whether it’s taxing or not, is undercutting my ability to be a moral actor".

In statements released by the American Humanist Association, Durman also pointed to broader objections regarding generative AI and large language models, citing concerns over worker livelihoods, environmental costs from data centers, and the displacement of human creativity.

Following a collegial mid-September meeting between Durman, Huiskens, and university officials to evaluate the sincerity of his beliefs, the university's Office of Equal Opportunity approved the request on September 25, 2026. Because Durman's primary role involves helping university researchers prepare grant applications—a task he demonstrated he could perform entirely on his own—the university determined that excusing him from personal AI use posed no "undue burden" on his department.

Section Summary:

  • Who and Where: David Durman, a 45-year-old research administrator at the University of Kentucky.
  • The Tools Cited: Microsoft Teams Copilot, ChatGPT, and Claude.
  • The Timeline: Requested in late August 2026; approved internally by the university's Office of Equal Opportunity on September 25, 2026.
  • The Trigger: Workplace encouragement to adopt AI tools rather than an existing mandatory policy.

The Operational Catch: Personal Use vs. Incidental Exposure​

If you are a systems administrator wondering how on earth an organization is supposed to quarantine a single user from every AI-backed system on a modern enterprise network, take a breath. The University of Kentucky's approval letter drew a sharp, pragmatic boundary that IT leaders should study closely.

According to Religion News Service, the senior manager in the university's OEO office explicitly wrote in the September 25 approval letter: "David may still be expected to participate in such meetings or be required to submit materials through such sites. David is excused from personal use of AI to complete his tasks."

In practical terms, Durman is not living in an analog bubble:

  • Meetings and Collaboration: He must still attend virtual or in-person meetings where AI tools—such as Microsoft Teams Copilot transcription or meeting summarization—are active.
  • Colleague Workflows: He must cooperate normally with coworkers who choose to use generative AI in their own tasks.
  • System Portals: He can still be required to upload or submit grant materials through university or third-party web portals that utilize underlying AI processing.

This distinction between personal cognitive delegation and incidental system exposure is the exact reason the accommodation worked without breaking departmental operations. Durman himself emphasized to Fortune and Religion News Service that his exemption only applies to his own task execution and that he has no desire to restrict how his colleagues work.

Workplace ActivityCovered by Durman's Exemption?Operational Reality for IT & Management
Prompting Microsoft Copilot, ChatGPT, or Claude to draft or summarize grant materialsYes (Excused)Employee completes research administration and writing tasks manually.
Attending a Microsoft Teams meeting where colleagues enable Copilot featuresNo (Required)Employee must still participate in meetings where AI tools are active.
Collaborating with coworkers who use AI in their own workflowsNo (Required)Colleagues remain free to use approved enterprise AI tools.
Uploading documents to institutional portals that use backend AI processingNo (Required)Standard software pipelines and submission portals remain unchanged.

Why "First-of-Its-Kind" Needs a Fact-Check​

While several outlets hailed Durman's case as the first religious exemption from workplace AI ever granted, the American Humanist Association was careful to define it more narrowly: it is the first known accommodation of its kind for a Humanist.

In fact, Durman told reporters he was directly inspired to seek help from the AHA after reading about Erin Maus, a software engineer and practicing Unitarian Universalist in North Carolina. As reported by Fortune and Religion News Service, Maus successfully obtained a religious accommodation from her tech employer in May 2026 after raising ethical and environmental objections to using AI in her software engineering role.

Furthermore, ethical friction around artificial intelligence is hardly confined to nontheistic philosophies or progressive congregations. Earlier in 2026, Pope Leo XIV issued an encyclical titled Magnifica Humanitas, warning that unchecked reliance on AI chatbots and automated systems could erode human dignity, critical thinking, and democracy while concentrating opaque power in the hands of a few elites. As religious and philosophical leaders across the spectrum formalize their stances on machine intelligence, HR departments are likely to see moral objections cited from a wide variety of traditions.

The Title VII Standard: Why Most Workers Can't Just "Opt Out"​

Does Durman's success mean any Windows user who despises the Copilot icon on their taskbar can claim a philosophical exemption on Monday morning? Not quite.

Workplace religious accommodations in the United States are governed by Title VII of the Civil Rights Act of 1964, enforced by the U.S. Equal Employment Opportunity Commission (EEOC). Under established federal law and EEOC guidance, Title VII protections are not limited to traditional, organized theistic religions like Christianity, Judaism, Islam, Buddhism, or Hinduism. Federal courts have long recognized that sincerely held moral or ethical beliefs—including secular Humanism—qualify for protection if they occupy a place in the believer's life parallel to that filled by a traditional faith.

However, an employee requesting an AI exemption must clear two substantial hurdles:

1. Sincerity vs. Personal Preference​

As constitutional and religious liberty law professor Douglas Laycock noted to Religion News Service, Title VII accommodations hinge on a fact-specific inquiry into whether an employee's belief is a deeply held religious or ethical commitment rather than a secular political, economic, or personal preference. According to Fortune, Durman was able to demonstrate a 30-year adherence to Humanism dating back to his teenage studies of Enlightenment philosopher David Hume and the Stoics, combined with a consistent personal habit of disabling AI features like Google's Gemini in his private life. An employee who simply dislikes AI hallucinations or worries about corporate layoffs will not meet the EEOC threshold.

2. The "Undue Hardship" Standard Under Groff v. DeJoy​

Even when a belief is undeniably sincere, an employer is not required to grant an accommodation if doing so imposes an "undue hardship" on the business.

For decades, employers relied on a low threshold where anything more than a de minimis (trivial) cost justified a denial. That changed with the U.S. Supreme Court's landmark 2023 ruling in Groff v. DeJoy (involving a Christian postal worker who objected to Sunday shifts). Under Groff and current EEOC policy, an employer must show that an accommodation would result in a burden that is "substantial in the overall context of an employer's business," taking into account the company's size, operating costs, and the practical impact on efficiency and other workers.

In Durman's case, writing and reviewing grant applications manually did not slow down his department or force coworkers to shoulder extra work. By contrast, if an enterprise hires a customer support specialist specifically to manage an AI-driven ticketing queue, or a developer whose core job is maintaining Azure AI Foundry pipelines, excusing that worker from touching AI would fundamentally alter the job requirements and almost certainly constitute an undue hardship.

Practical Takeaways for IT Admins and Enterprise Leaders​

From an IT governance perspective, the University of Kentucky case is a textbook example of why software deployment cannot be treated purely as a licensing exercise. As organizations roll out Microsoft 365 Copilot, Teams Premium features, and third-party LLM integrations, technical capabilities are colliding directly with human resources policy.

If you help manage enterprise IT or workplace technology policies, here is how to prepare for the inevitable accommodation requests:

  • Maintain Granular Licensing and Feature Controls: In Microsoft 365 admin centers, Copilot licenses and Teams AI features can—and should—be assignable at the individual user or security-group level rather than forced globally without exception switches. If HR grants an employee a personal-use exemption, IT should be able to cleanly revoke or hide their direct Copilot seat without disrupting tenant-wide compliance or meeting transcription for the rest of the team.
  • Define Whether AI is an "Essential Job Function": Before mandating AI tools across a department, leadership and HR should document whether using a specific AI assistant is the essential outcome of the role or merely an optional method for achieving standard productivity metrics. If an employee can hit all key performance indicators manually without burdening peers, denying an accommodation becomes much harder under the Groff standard.
  • Separate Direct Use from Backend Infrastructure: Follow the University of Kentucky’s blueprint by distinguishing between active cognitive prompting (asking an LLM to write a report) and passive system interaction (attending a recorded Teams call or submitting data to an automated portal). Excusing a user from typing prompts into a chatbot is administratively simple; excusing them from interacting with enterprise databases that use machine learning under the hood is practically impossible.

(Note: The administrative breakdown above is a synthesis of enterprise IT management practices and public EEOC guidance, not formal legal advice. Organizations facing Title VII requests should always consult qualified employment counsel.)

The Bigger Picture for the AI-Driven Workplace​

David Durman's stand in Kentucky highlights a growing cultural tension inside the modern office. While tech executives—from Shopify CEO Tobi Lütke calling AI usage a baseline expectation to Amazon urging engineers to adopt automated coding tools—press forward with aggressive deployment targets, a significant slice of the workforce remains deeply skeptical. Whether driven by Pope Leo XIV's warnings on human dignity, Unitarian Universalist environmental ethics, or a secular Humanist commitment to doing one's own thinking, employees are testing where the machine ends and human agency begins.

For now, Durman gets to keep drafting and reviewing university grant proposals the old-fashioned way: using his own brain, while Microsoft Copilot hums along quietly on his coworkers' screens down the hall.


Update: Additional details (October 31, 2026)​

University of Kentucky spokesperson Jay Blanton told TechCrunch that accommodations may be reviewed as technology and job responsibilities change. He said the university determined Durman could perform his essential job functions with the accommodation in place.

 

References

  1. Man wins first-of-its-kind religious exemption to avoid using AI at work - Dexerto Dexerto 2026-10-30T14:44:43+00:00
  2. Man wins religious exemption from AI, a first for humanists What's Trending · 2026-10-31T01:16:08+00:00
  3. As bosses push workers to use AI, a Kentucky man found a way to opt out | Fortune fortune.com