"AI detection is very unlike plagiarism"
Thousands of UK university students have been caught misusing ChatGPT and other AI tools, as traditional plagiarism sharply falls.
According to a Guardian investigation, there were nearly 7,000 confirmed cases of AI-related cheating in 2023-24. That equates to 5.1 cases per 1,000 students, more than triple the rate of the previous academic year.
Early figures from 2025 suggest the trend is accelerating, with about 7.5 proven cases per 1,000 students projected. But researchers say these figures barely scratch the surface.
Dr Peter Scarfe, an associate professor of psychology at the University of Reading, warned:
“I would imagine those caught represent the tip of the iceberg.
“AI detection is very unlike plagiarism, where you can confirm the copied text. As a result, in a situation where you suspect the use of AI, it is near impossible to prove, regardless of the percentage AI that your AI detector says (if you use one).
“This is coupled with not wanting to falsely accuse students.
“It is unfeasible to simply move every single assessment a student takes to in-person.
“Yet at the same time the sector has to acknowledge that students will be using AI even if asked not to and go undetected.”
The shift to AI marks a significant change in how academic dishonesty occurs.
In 2019–20, plagiarism made up nearly two-thirds of all academic misconduct. During the pandemic, plagiarism increased further as assessments moved online.
But the rise of generative AI tools is now reshaping the landscape.
Confirmed plagiarism cases fell from 19 per 1,000 students in 2019–20 to 15.2 in 2023–24. Preliminary data suggests this could fall further to just 8.5 per 1,000 students this year.
One hundred and thirty-one universities provided some data, though not all universities had complete records for each year or category of misconduct.
Over a quarter of respondents (27%) did not yet record AI misuse separately, suggesting many universities are still adjusting to the new challenges.
Meanwhile, actual usage of AI tools may be far higher.
A February survey by the Higher Education Policy Institute found 88% of students had used AI in assessments.
Researchers at the University of Reading tested their own assessment systems and found they could submit AI-generated work without detection 94% of the time.
Social media platforms have also fuelled the issue.
Dozens of TikTok videos promoting tools that “humanise” ChatGPT text to help students evade detection have been uncovered.
Dr Thomas Lancaster, an academic integrity researcher at Imperial College London, said:
“When used well and by a student who knows how to edit the output, AI misuse is very hard to prove. My hope is that students are still learning through this process.
“University-level assessment can sometimes seem pointless to students, even if we as educators have good reason for setting this.
“This all comes down to helping students to understand why they are required to complete certain tasks and engaging them more actively in the assessment design process.
“There’s often a suggestion that we should use more exams in place of written assessments, but the value of rote learning and retained knowledge continues to decrease every year.
“I think it’s important that we focus on skills that can’t easily be replaced by AI, such as communication skills, people skills, and giving students the confidence to engage with emerging technology and to succeed in the workplace.”
Commercial tech providers are increasingly targeting students. Google offers a free upgrade to its Gemini AI tool for university students. OpenAI gives discounts to students in North America.
Science and Technology Secretary Peter Kyle said AI should be used to “level up” opportunities for dyslexic children.
A government spokesperson said it was investing over £187 million in national skills initiatives and had published AI guidance for schools.
They said: “Generative AI has great potential to transform education and provides exciting opportunities for growth through our plan for change.
“However, integrating AI into teaching, learning and assessment will require careful consideration and universities must determine how to harness the benefits and mitigate the risks to prepare students for the jobs of the future.”
Loading ...