Yale faces federal lawsuit over AI-cheating dispute

A legal battle between Yale University and a former Executive MBA student who paid $208,500 in tuition has escalated into a 13-count federal lawsuit that began in February 2025. The dispute arose after the student was accused of using generative AI on a four-hour final exam for the course MGT423E, Sourcing and Managing Funds. Out of 72 students, only his exam was flagged, prompting a professor to run the text through the AI detector GPTZero on June 11, which flagged several sections, leading to an incomplete grade on June 12. The school gave the student an F and suspended him for a year, leading to a sprawling legal case that has accumulated 125 docket entries.

The university's investigation focused on whether the student used ChatGPT to write his exam, noting he performed poorly on question 5 where AI was least helpful. Administrators repeatedly requested the original file to verify its metadata. On August 12, August 16, and August 19, investigators pressed him for a Microsoft Word file. The student defended himself by scanning documents published over 30 years ago to prove the detector's inaccuracy, but during a November 8 hearing, he revealed he actually used Apple Pages. Following the hearing, events moved rapidly: the session ended around 2 pm, the student sent the Pages file at 3 pm, administrators called at 3:30 pm and texted at 4 pm, the student texted at 5:07 pm declining a 5:30 pm meeting, and by 8:28 pm, he was penalized for what the university termed not being forthcoming. The legal battle has dragged on, with a judge warning the student in June 2026 over his third amended complaint, followed by a Yale motion to dismiss on July 15, 2026.

For AI practitioners and academic administrators, this case underscores the severe risks of relying on automated detection tools like GPTZero, which are notoriously unreliable and prone to bias against non-native English speakers. It demonstrates that institutions cannot rely solely on algorithmic flags to police academic integrity without establishing clear, legally defensible forensic protocols. Practitioners must realize that metadata preservation, clear communication about file formats, and transparent investigation timelines are critical to avoiding protracted, multi-year litigation when AI use is suspected.