AI Cheating Dispute at Yale Becomes Federal Case

Thierry Rignol is furious with Yale. After shelling out a fortune in tuition for the Executive MBA, he was accused of cheating and suspended for a year, in addition to receiving an F in the Sourcing and Managing Funds course. This, he claims, despite being one of the top students in the class. The central issue? Artificial intelligence.

Rignol's final exam was flagged as possibly written with the help of AI. The professor used a detection tool called GPTZero, which indicated that parts of the text appeared to be generated by AI. But these tools are notoriously flawed. Even Yale admits that it is difficult to control the use of AI solely with detectors. Rignol argues that his long, well-formatted responses are the result of his own intelligence. He also points out the tool's bias against non-native English speakers, like himself.

Rignol believes that Yale is against him, censoring his conservative political speech. The case now includes 13 allegations against Yale, from breach of contract to defamation. He seeks "unlimited" damages to cover economic losses and reputational harm. Despite everything, Rignol still wants to be readmitted and have his F overturned.

Yale and the Mystery File Issue

The story began with the final exam of spring 2024. Of the 72 students, only Rignol's exam was flagged for possible AI use, partly due to its unusual length. Professor K. Geert Rouwenhorst reported several concerns, including similarities to responses generated by ChatGPT. Additionally, Rignol performed poorly on a question where AI would be less helpful. Yale questions whether anyone could have produced such an extensive and polished exam in just four hours.

During the summer of 2024, the investigation continued. Rignol claimed innocence, stating that the exam reflected his exceptional performance and distinct writing style. He felt that Yale was pressuring him to falsely confess, even mentioning the fear of deportation. Yale, on the other hand, sees Rignol as evasive. Professor James Choi, who led the investigation, repeatedly requested the original file used by Rignol to draft his responses, but he never provided the document.

The Case Continues Without Resolution

In August, Choi directly asked Rignol for the Word file that generated the PDF of the exam. He reminded him that a lack of cooperation could result in permanent expulsion. Rignol, however, did not submit the document. In September, the Honor Committee officially notified Rignol about the allegation of improper use of AI. A meeting was scheduled for October, but Rignol requested a postponement due to his wedding in Europe.

Back in November, days before the hearing, Rignol questioned the composition of the committee, arguing that student members should be recused if they had not taken the same courses and exercises as he had. Yale's response was that there was no basis for such recusals.

The case continues to drag on, with no signs of resolution. The central question remains: Was Rignol a victim of a flawed algorithm and political hostility, or is there more to this story? Yale seems to believe in the latter option.