These four Russell Group universities actually expelled students for using AI

Over 2000 students were flagged for AI-related cheating last year

It feels like we can’t go a day without hearing another debate about the place of AI in our education system, especially in regard to university work.

As the general use of artificial intelligence in our day-to-day lives increases, it seems inevitable that its use in academic work also goes up. But, in cases where university rules are breached, students using AI have faced large-scale punishment. 

Data published by The Times reveals over 2000 students have been punished for using generative AI in their submitted university work in the 2024/25 academic year, compared to around 700 students in 2023/24.

According to data released in response to Freedom of Information (FOI) requests, four Russell Group universities were found to have expelled students due to findings of AI-related cheating, though it is suspected the actual rate of cheating through AI use may be significantly higher than discovered.

A Russell Group professor claimed “most” students were using AI in their work, but that only the “most egregious” cases were ever caught and punished. The professor argued “the more devious students” are aware of aspects of their text that signal the use of AI, and “edit accordingly” to avoid being caught and punished.

This, however, has resulted in many students who do not use AI to change their writing styles, including the infamous death of the em dash as AI-detection software flags it as generative AI use. Many universities use software such as Turnitin for assessed work, but this is frequently criticised among student bodies for flagging student-written work.

These Russell Group universities expelled students for AI-related cheating

Glasgow University building

via Unsplash

  1. University of Glasgow
  2. University College London (UCL)
  3. University of Leeds
  4. Imperial College London

In the 2024/25 academic year, there were 2,053 recorded punishments for AI use from a sample of 350,000 students, which equates to a rate of approximately one in 170 students. Seven Russell Group universities, however, did say they do not record AI-related investigations or sanctions.

Hollie Chandler, the Director of Policy at the Russell Group, emphasised that students who were caught using AI represent “a very small proportion of the student population”.

In a survey conducted this year, Higher Education Policy Institute (HEPI) highlighted that 12 per cent of British students “admitted copying AI-generated text directly into assessments,” a significant increase from three per cent of students in 2023. 94 per cent of students admitted that they utilise AI to help with assessed work in some way.

For more like this, like The Tab on Facebook.  

Featured images via Unsplash

More on: AI Russell Group University