Student fined for AI-generated appeal

The administrative court in Rennes has fined a student named Delphine 750 euros for submitting an appeal it deemed abusive. The student had been rejected from a master's program in psychology at the University of Rennes 2 and filed an appeal against the decision.

Appeal flagged for AI characteristics

The court ruled that the appeal showed signs of generative artificial intelligence. These included a mix of curved and straight apostrophes, frequent use of em dashes, and round black bullets. The judge noted the text was overly orderly, pedagogical, and repetitive, traits often seen in AI-generated content.

Late appeal and lack of urgency

The student had been rejected on June 3, 2026, but filed the appeal on October 18, 2026. She cited a busy summer and administrative tasks as reasons for the delay. However, the court found no urgency or serious doubt about the rejection's legality.

Student's arguments dismissed

The student argued the university lacked clear admission criteria and had irregularities in the rejection process. The court found these claims unconvincing, noting the absence of documents on the university’s website did not prove their non-existence. The student’s file also did not support her claims.

AI tool cited and appeal errors noted

The court mentioned Claude.ai, a chatbot developed by Anthropic, as a possible tool used. It also pointed out errors in the appeal, such as incorrect references to "Madame B." and "Madame C." A similar appeal had been recorded previously, suggesting the text may have been generated in bulk without proper updates.