🔍 Read the full analysis: Where US Workplace Law Protects—and Limits—AI Refusals on ThorstenMeyerAI.com
Get business pricing on office and shipping supplies
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
TL;DR
The University of Kentucky approved research administrator David Durman’s request to be excused from personally using AI tools, finding that he could still perform his job without them. The decision is an individual workplace accommodation, not a court ruling or legal precedent, and it does not exempt him from every workplace system that uses AI.
The University of Kentucky approved research administrator David Durman’s request to be excused from personally using artificial intelligence tools at work, an individual accommodation based on his sincerely held Humanist beliefs. The September 25 decision allows him to continue his job without personally using AI, but it is not a court ruling and does not establish a legal precedent for other workers.
Durman helps researchers apply for grants. In a letter dated August 24, he asked the university’s Office of Equal Opportunity to excuse him from personally using AI and large language model tools, including automated transcription, meeting notes and decision-making systems. He offered to take and share notes himself. The American Humanist Association (AHA) assisted with the request.
The university approved it on September 25, after a meeting between Durman, AHA staff attorney Alexandria Huiskens and university officials. The AHA said the office found the accommodation would pose “no undue burden on his department.” A university spokesperson told TechCrunch the arrangement was possible because Durman could still perform the essential functions of his position.
The approval is limited in scope. According to the letter as described by Religion News Service, Durman may still need to attend meetings where AI is used and submit materials through systems that use it. The decision concerns his personal use of AI, not a blanket exemption from all AI-related workplace processes. The university has said accommodations may be reviewed as jobs change.
Infographic 01 · University of Kentucky · 2026
One letter, one meeting, one narrow exemption.
A Humanist research administrator asked not to use AI himself. His employer, which rolled out Copilot campus-wide, said yes.
- 31 Jul 2026Copilot promoted for meeting notesAt a department meeting, per Durman’s letter. Chatbots suggested, internal AI tools in development.
- 24 AugTwo-page letter to the Office of Equal OpportunityRequest under Title VII, drafted with the American Humanist Association. “Using AI or LLM tools is not an essential function of my position.”
- Mid-SepOne meetingDurman and the AHA’s staff attorney meet university officials. “A very collegial meeting.”
- 25 SepApproved“No undue burden on his department,” per the AHA. No lawsuit, no court, no precedent.
- 26 to 29 OctMade publicAHA, Religion News Service, then TechCrunch.
Excused
- Personal use of AI or LLM tools to complete his own tasks
- AI note-taking: he takes and shares his own notes
Still expected
- Taking part in meetings that use AI
- Submitting materials through systems that use AI
- Same duties; colleagues keep using AI
- May be reviewed “as technology and job responsibilities evolve”
“David is excused from personal use of AI to complete his tasks.”University approval letter, as quoted by Religion News Service. The letter itself is not public.
Sources: Durman letter, 24 Aug 2026; AHA press release, 26 Oct 2026; RNS, 26 Oct 2026 (quotes the approval letter); TechCrunch, 29 Oct 2026 (university statement). Timeline and layout: Thorsten Meyer.
Powered by Thorsten Meyer AI
The Limits of an AI Opt-Out
The case shows how existing workplace protections may apply when an employee’s religious or ethical convictions conflict with a workplace rule or expected tool. Under Title VII of the Civil Rights Act, covered employers generally must reasonably accommodate an employee’s sincerely held religious belief when it conflicts with a work requirement, unless doing so would cause undue hardship. The law covers employers with 15 or more employees.
That does not give every worker an automatic right to refuse AI. The employer may consider the employee’s duties, the tool’s role in the job and the practical effect of an accommodation. The university’s stated basis was that Durman could continue carrying out his essential work without personally using AI. A job in which use of an AI system is central could present different facts and a different accommodation analysis.
The case also matters because employers are introducing AI into routine tasks such as meeting records and document handling. It illustrates that a workplace’s broader adoption of a tool does not, by itself, answer whether an individual accommodation is workable. The decision remains specific to one employee and one employer.
Products Worth Considering
Logitech MX Master 4 for Mac, Ergonomic Bluetooth Mouse with Haptic Feedback, Ultra-Fast Scrolling, USB-C Charging, for MacOS, iPadOS, Windo
Dell 34-Inch S3425DW Curved WQHD USB-C Monitor
Dell XPS 13 9350 Laptop, 13.4 FHD+ Display, Intel Ultra 7 256V, 16GB DDR5, 1TB SSD
As an affiliate, we earn on qualifying purchases.
How the Kentucky Request Developed
According to Durman’s accommodation letter, AI note-taking was promoted at a department meeting on July 31. The letter referred to Microsoft Teams Copilot, commercial tools such as ChatGPT, Claude and Gemini, and internal AI systems under development. Durman said he was asking in anticipation of possible future requirements; TechCrunch reported that he was not yet required to use the tools.
In his August 24 request, Durman described himself as a Humanist who does not believe in a god. Federal employment guidance does not limit religion to belief in a deity. The Equal Employment Opportunity Commission’s regulation includes moral or ethical beliefs about right and wrong when they are sincerely held with the strength of traditional religious views. A personal preference or political position, on its own, is not necessarily covered.
The university is an AI-adopting institution. Microsoft has described its deployment of Microsoft 365 Copilot as reaching more than 70,000 students and employees; that figure is Microsoft’s own account. Earlier, Business Insider reported that software engineer Erin Maus, a Unitarian Universalist in North Carolina, received an exemption from using AI in May 2026. The AHA calls Durman the first Humanist to receive one; that distinction is the organization’s claim, not a court finding.
“I believe that our capacity to think and reason for ourselves is worth protecting.”
— David Durman, in a statement cited by the American Humanist Association
What the Approval Does Not Decide
The accommodation does not resolve whether another employee could refuse an AI tool, or whether a different employer would have to grant a similar request. Those questions would depend on the worker’s belief, duties, the requested adjustment and its effect on the employer. No lawsuit or court ruling was reported in this case, so it creates no binding legal precedent.
The available reporting also leaves practical details open, including how the university will handle future changes to Durman’s role or systems he must use. The reported terms distinguish personal use from participation in meetings or workflows involving AI, but the full day-to-day application may depend on how those tasks are arranged. The AHA’s claim that Durman is the first Humanist to receive such an accommodation has not been independently established as a legal finding.
How the Accommodation May Change
The university’s stated position is that accommodations can be reviewed as jobs change. The next relevant development would be any change to Durman’s duties or to the AI systems used in his department, and whether the existing arrangement still allows him to perform the essential functions of his role. The university has not announced a timetable for a review.
For other employees, the case points to an individual process rather than a general AI opt-out rule. A worker considering a request would need to explain the conflict and the adjustment sought; the employer would assess whether the belief qualifies for protection and whether an accommodation is workable. Under the Supreme Court’s 2023 decision in Groff v. DeJoy, an employer claiming undue hardship must show substantial increased costs in relation to the conduct of its particular business. The outcome in any other case remains fact-specific.
Key Questions
Can a US employee refuse to use AI at work?
Sometimes, but there is no general right to refuse every workplace AI tool. Under Title VII, a covered employer may have to accommodate a sincerely held religious belief that conflicts with a work requirement, unless the accommodation would cause undue hardship. The result depends on the facts of the job and request.
Does a religious accommodation require belief in God?
No. EEOC guidance recognizes that qualifying beliefs can include nontheistic systems and sincerely held moral or ethical beliefs. A personal preference or political opinion alone does not automatically qualify.
What did the University of Kentucky approve?
The university excused David Durman from personally using AI tools to do his work. Reporting on the approval says he may still attend meetings involving AI or use systems that incorporate it, so the arrangement is not a complete exemption from AI-related processes.
Does Durman’s case set a legal precedent?
No. The University of Kentucky granted an individual workplace accommodation. There was no lawsuit or court decision, and the approval does not bind other employers.
What can count as undue hardship?
Following the Supreme Court’s 2023 Groff v. DeJoy decision, an employer denying a religious accommodation must show substantial increased costs in relation to the conduct of its particular business. Whether that standard is met depends on the circumstances.
Source: ThorstenMeyerAI.com
