A professor at a university institute in France decided not to correct the exams of her 120 first-year students because of excessive use by students of the artificial intelligence software ChatGPT.
“This is a class of about 120 freshmen […] who had to submit a written document on hydrotherapy”, explained Monday Samuel Cruz-Lara, the director of the IUT Charlemagne in Nancy (east).
“When grading the marketing assignments, the teachers in charge of the correction realized that the students had made excessive use of ChatGPT: their copies did not include the slightest spelling error, or were composed of searched phrases” , continued Mr. Cruz-Lara.
“If we had had more time, we could have asked the students to explain themselves, which we did with some, who recognized things half-wordly. Except that we are at the end of the academic year and in terms of the calendar it is complicated for us to react.
The professor responsible for the subject decided collectively, with her other corrector colleagues who intervene in the module, not to correct the disputed written test.
The director of the IUT, however, wanted to be reassuring: “The students will still have a mark for this subject, they will not be penalized.”
For the future, he decided “not to cut off access to this kind of tool from the IUT”, but to “engage with the students in a benefit-danger discussion and to support them in the use of ChatGPT”.
A modification of the internal regulations of the IUT is also envisaged: “We already had a paragraph on plagiarism, I will ask to add one concerning artificial intelligence type tools, the use of which will be sanctioned in the same way as plagiarism. .”
For its part, the University of Lorraine, of which the IUT Charlemagne is a component, “has not yet taken a general decision” on the use of artificial intelligence.