Fear of being flagged by AI detectors drives stress among students

Universities have been urged to reconsider their use of tools that claim to be able to detect artificial intelligence after a survey found three-quarters of UK students using AI feel stressed that their work will be wrongly flagged as cheating.

Polling by YouGov, commissioned by student support company Studiosity, found that AI use is continuing to rise in universities as more older students and women join early adopters in utilising the technology.

But this has been accompanied by increased stress among students, many of whom are still unclear what is and isn’t acceptable use, with the risk of being accused of wrongdoing fuelling much of this anxiety.

The survey of 2,373 students found 60 per cent experienced stress while using the tools, with international students twice as likely to say they experienced “a lot” of stress.

More than half (52 per cent) of all those polled cited “being accused of cheating when I did nothing wrong” as one of the factors causing them stress. Of the AI-using students, 75 per cent reported significant stress over being wrongly flagged for plagiarism by detection tools.

Those who said they had only used the tools once or twice or a few times were more likely to feel stressed about being accused of cheating than those who said they used them regularly. 

Further factors causing stress include becoming addicted to AI (40 per cent) and concerns about ownership of the work (40 per cent).

“This highlights a significant trust gap between students, their tools, and institutional detection methods,” the report says.

It recommends that university leaders should “reconsider detection tools that give false positives” in light of the findings and “establish pathways to protect students from wrongful ‘cheating’ accusations”.

Overall, 71 per cent of respondents reported using AI tools to help with assignments or study tasks, up from 64 per cent when similar polling was conducted last year.

The increase came from more students aged over 26 adopting AI tools, with 76 per cent now saying that they have, compared with 66 per cent previously.

There was also a significant increase in adoption among female students, narrowing the gender gap, although male students still used AI more frequently.

Usage among domestic students has also grown 9 per cent, although the 69 per cent of these students who now use AI still lags behind the 87 per cent of international students who do.

The survey also found AI use was very high and increasing among law and business students (75 per cent and 80 per cent respectively) but less widespread among those studying humanities and social sciences (58 per cent) and creative arts 52 per cent).

Institutional responses to AI still vary widely, with some students reporting that “traffic light” systems were used to outline what is and isn’t acceptable.

But other institutions were said to operate under more restrictive environments, relying on blanket bans and detection tools.

Asked if they would rely entirely on AI to write their papers if permitted to do so, only 21 per cent of students said they would, with almost half expressing concerns about how AI is eroding their critical thinking and communication skills.

Writing in the foreword to the report, Universities UK chief executive Vivienne Stern said: “It is important that students have confidence in using AI to enhance, rather than replace, genuine learning and that they are clear about what is and is not permitted by their university. This report shows we still have some way to go in this regard.”

A high percentage of students still don’t feel they are receiving enough support, Stern says, which could “lead to anxiety about what is and is not permitted”.

“While universities are learning too, this confidence and clarity gap needs to be narrowed and eventually closed.”

tom.williams@timeshighereducation.com