evidoria

← Back to browse

Good practice Imported

Yale's GPTZero Wrongful-Suspension Lawsuit — An MBA Student Fights a Year-Long Ban Over an AI-Detection Flag

United States of America · New Haven · See the United States of America profile

Evidence: Descriptive / self-reported Top 98% 13/100 · Ask Evidence Copilot about this practice

A Yale MBA student was suspended for a year after GPTZero flagged his no-internet exam as AI-written; his federal lawsuit alleges national-origin discrimination and due-process violations, becoming a cautionary test case for AI-detector-based discipline.

Details

Promoter
Yale School of Management
Period
June 2024 (flagged) – ongoing 2025–2026 (federal lawsuit)
Keywords
Higher education, academic integrity, AI detection, civil rights litigation

Context

In June 2024, a teaching assistant in Yale School of Management's 'Sourcing and Managing Funds' course flagged the open-book, no-internet final exam of MBA student Thierry Rignol, a French national, as likely AI-written, citing 'unusually long and elaborate' formatting and 'near-perfect punctuation and grammar', and ran portions of the answers through the AI-detection tool GPTZero to support the suspicion.

Activities

Yale's Honor Committee found Rignol liable both for the original AI-use allegation and for 'not being forthcoming' during the investigation, imposing a one-year suspension from the Executive MBA programme and a failing grade in the course.

Results

In February 2025, Rignol filed a federal lawsuit against Yale and five named officials, alleging national-origin discrimination, due-process violations, coercive investigative tactics, breach of contract, and defamation. A judge denied Rignol's request for a preliminary injunction, and litigation continued into 2026 with a jury trial sought.

Conclusions

Yale's own materials acknowledge that no AI-detection tool can establish AI use with certainty. The case has become a widely cited cautionary example of the legal and equity risks of disciplining students based on unreliable AI-detection tools, echoing a 2025 FTC action against an AI-detector vendor whose advertised '98% accuracy' was found to perform near chance -- findings that have led several universities to disable AI-detection software altogether.

Implementation
Indicative cost
Low (< €50k) — GPTZero is a low-cost SaaS subscription; litigation costs not disclosed.
Time to results
Medium (1–3 years) — Exam flagged June 2024; suspension imposed; federal lawsuit filed February 2025; ongoing into 2026.
Staffing & skills
Yale School of Management Honor Committee, GPTZero (AI-detection vendor)

Common failure modes

  • disciplinary action based substantially on an AI-detection tool whose vendor and the university both acknowledge cannot establish AI use with certainty
  • led to a federal discrimination and due-process lawsuit still unresolved as of 2026

Where it fits

Governance type
university academic-integrity process
Scale
single institution, single case with broader precedent value
Income level
high income (United States)

Commonly funded by

Own resources / municipal budget

Indicative funding routes for practices of this type — always check each programme's current calls and eligibility rules.

Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.

Data sources

Where this practice's information was retrieved from, and when.

Similar practices you may find useful