According to Marike Polak, chair of the examination board of the Erasmus School of Social and Behavioural Studies (ESSB), both cases are minor incidents. “The vast majority of fraud cases perpetrated by students are minor infringements, often committed due to personal circumstances or time pressure”, says Polak.
Resit
Similarly, these two cases probably did not involve the copying of large sections of text. “In such a case, students are often allowed a resit, provided that they did not commit fraud before and have written a reflective essay on the incident assessed to be satisfactory by the examination board”, says Polak. She cannot give any details on the specific cases, as they are still under investigation.
The faculty recently sent all its students an email to inform them of the two possible cases of fraud. Polak: “The email was intended as a warning, but we don’t want to create the impression that we have started a witch hunt.”
According to a spokesperson for the university, there have been no reports or known cases of this type of fraud in other faculties.
Own texts
Fraud involving ChatGPT can be identified because text is similar to published text from academic articles, says Polak. In addition, when text from articles is incorporated incorrectly, such as when the content of the running text doesn’t match with the cited academic text, this may also indicate fraud involving use of the bot.
In line with all the university’s other examination boards, the examination board of the ESSB considers the copying of AI-generated texts to be fraud, as students should always personally write their own texts.
Search engine
Students of the ESSB are allowed to use this technology as a search engine, says Polak. But they should always check the information provided by the bot against primary sources, because the bot doesn’t always give the right answers. In addition, students should personally look up sources, make analyses and write down the information in their own words.
