A leading university plans to grill students on work they submit, to put them off cheating by using chatbots. New artificial intelligence software – such as Microsoft-backed ChatGPT – is able to compose instant, “human-like” and accurate responses to any question it is asked.
Experts predict it will cause an explosion of dishonest students. But now undergraduates at Imperial College London are to be randomly selected by academics for an oral examination about their completed assessments.
ICL is the first university to put such measures in place – even where there is no specific concern the work might not be the student’s own.
Lecturers will question them “about the subject and how they approached their assignment”.
It is hoped the fear of an interview will act as a deterrent.
Large numbers of university degree courses are assessed by coursework and essays, so are vulnerable to cheating.
Where exams are still common, many have moved online, despite evidence that it leads to more misconduct.
Version:1.0 StartHTML:000000096 EndHTML:000001518 StartFragment:000000096 EndFragment:000001518
Critics have called for universities to reinstate in-person exams or put in place other measures to protect academic integrity.
Chris McGovern, the chairman of the Campaign for Real Education, said: “The move by Imperial College is placing the integrity of academic qualifications above cheating. All universities should follow their example. They should also expel cheats and cancel their degrees.”
Meanwhile, schools are also grappling with how to deal with AI use among pupils.
Teachers and lecturers have found the text generated by ChatGPT and Google rival Bard are already of a high enough standard to pass various tests.
Advice issued by exam boards last week says teachers should supervise all course work that counts towards GCSEs and A-levels by timetabling it within normal classroom hours.
Imperial College London was contacted for comment.
