Only 25 per cent of college lecturers can detect if college students’ work is AI generated

Only one in four university lecturers feel confident they can detect work that has been generated by artificial intelligence, a survey has found.

Students reported using AI for nearly half (48 per cent) of their studies, with four in five (80 per cent) saying it has improved their grades.

But the number of professors and lecturers who feel confident spotting AI-generated work has fallen from 42 per cent last year to 25 per cent, Coursera’s AI in Higher Education report found.

More than two in three (68 per cent) of the 500 university professors and lecturers surveyed by Censuswide said they use AI in their work at least often.

Only 27 per cent said lecturers at their university have the right skills to use AI effectively.

A majority of lecturers (55 per cent) and students (67 per cent) said rising AI usage has made passing exams easier.

Half of lecturers (50 per cent) and fewer than half of students (47 per cent) said using AI to help with university work should be considered cheating.

Just one in four university lecturers felt confident they could detect student work which was generated by artificial intelligence

Asked about the overall impact of AI on higher education, lecturers were more likely to say it has been negative (15 per cent) than students (6 per cent), with the majority of both saying it has had a positive impact.

Marni Baker Stein, chief content officer at Coursera, said AI is becoming ‘ubiquitous’ in higher education.

‘However, many educators are finding it harder to keep up with the practical and professional implications – contributing to a decline in confidence around skills and AI detection,’ she added.

More training will be needed to ensure lecturers feel confident spotting and using AI, she said.