Law student convicted of “serious fraud” with ChatGPT
In July, a law student was punished for “serious fraud” involving AI at the University of Amsterdam. The student had used ChatGPT, but it cited sources that did not exist. The examination board therefore excluded him from the course in question and two examinations in other courses.
A law student was convicted on appeal in a court case in July for using AI and submitting four non-existent sources. The UvA student had brought the case against his faculty’s examination board because he considered the penalty he had received to be disproportionately high. Earlier in the academic year, he had cheated using ChatGPT on an assignment for the Amsterdam Law Firm 2.1 course. As a result, the student was completely excluded from that course and from the examinations for two other courses.
In the autumn of 2024, the bachelor’s student had to write a paper entitled “Jurisprudence and Literature Research” with a bibliography together with his group members. However, when checking the paper, his seminar teacher noticed that sources 2 to 5 of the assignment did not exist at all. After receiving an email from the teacher, the student immediately provided links to four other, existing sources on the same day.
This led the seminar lecturer to suspect that the student had initially used artificial intelligence. The lecturer then reported him to the examination board. During a meeting with the board, the student admitted to having used ChatGPT. The chatbot had automatically filled in incorrect, non-existent sources.
Incorrect authors
The examination board referred to this as “serious fraud”. The members of the board held it against the student that he had also cast suspicion of fraud on his group members without telling them. They also considered it reprehensible that the use of AI had linked literature to authors who had not written it at all. The administrative court judge who assessed the case on appeal saw it no differently.
In the Amsterdam Law Firm course, law students learn to work in the legal world in practice. There are projects at large law firms such as Loyens & Loeff, Houthoff and NautaDutilh, and assignments at the Public Prosecution Service or the court. The aspiring legal professionals work in groups. This often requires them to examine complex legal issues, which necessitates research into previous court rulings. In one such assignment, the student who was punished used AI unlawfully. The use of AI that makes it impossible to assess a student’s skills is punishable under the examination regulations.
The law student is not the first UvA student to be convicted in court for using AI. Earlier this year, in July, a master’s student in constitutional and administrative law was tried for large-scale AI use in a writing assignment. Several dates, case numbers and substantive rulings did not match the sources provided. She was punished by the examination board with a six-month suspension from her studies.
