Thousands of university students across the UK have been caught misusing artificial intelligence tools such as ChatGPT to cheat on academic assignments, as traditional plagiarism rates continue to decline, according to a detailed investigation by The Guardian UK.
The report, based on data obtained under the Freedom of Information Act from 131 universities, reveals a dramatic surge in confirmed cases of academic misconduct involving AI tools. In the 2023–24 academic year alone, nearly 7,000 cases of AI-related cheating were recorded — equivalent to 5.1 for every 1,000 students, up from just 1.6 the year prior. Projections for the current year indicate that this figure could rise to 7.5 cases per 1,000 students.
By contrast, traditional plagiarism — once the dominant form of academic misconduct — is waning. Before the rise of generative AI, plagiarism made up nearly two-thirds of misconduct cases. But confirmed instances have dropped from 19 per 1,000 students in 2019–20 to 15.2 in 2023–24, with early indications suggesting this could fall to as low as 8.5 per 1,000 in the current academic cycle.
The investigation by The Guardian highlights the evolving nature of academic dishonesty, with students increasingly using tools like ChatGPT not to copy and paste ready-made content, but to generate ideas, structure assignments, and even rephrase AI output to evade detection.
Dr. Peter Scarfe, associate professor of psychology at the University of Reading, co-authored a study in which researchers successfully submitted AI-generated work that went undetected 94% of the time. “AI detection is very unlike plagiarism, where you can confirm the copied text,” Scarfe said. “In a situation where you suspect the use of AI, it is near impossible to prove.”
More than a quarter of the universities surveyed did not yet track AI misuse as a separate category, revealing how unprepared many institutions are for the technological shift. This gap is particularly concerning given a February survey by the Higher Education Policy Institute found that 88% of students admitted to using AI tools for assessments.
Dr. Thomas Lancaster, an academic integrity researcher at Imperial College London, noted that AI misuse is difficult to prove when students skillfully edit the output. “My hope is that students are still learning through this process,” he said, emphasizing the importance of integrating AI literacy into curricula rather than solely policing its misuse.
Students interviewed by The Guardian confirmed the widespread use of AI tools for brainstorming, structuring, and enhancing their work. One student, “Harvey,” explained, “ChatGPT kind of came along when I first joined uni… I don’t think many people use AI and then copy it word for word. It’s more just to help brainstorm and create ideas.”
Another student, “Amelia,” noted the value of AI tools for neurodiverse learners. “One of my friends uses it… to structure her own points. She has dyslexia – she said she really benefits from it.”
The growing normalization of AI in education is not limited to students. Technology companies such as Google and OpenAI are actively marketing AI tools to university students, offering free upgrades and discounts. Meanwhile, UK Science and Technology Secretary Peter Kyle told The Guardian that AI could help “level up” education for children with learning challenges.
Educators and policymakers are now grappling with how to adapt. Some have suggested a shift back to in-person exams, but others argue for a deeper rethink of assessment methods. “The value of rote learning and retained knowledge continues to decrease,” said Lancaster. “We should focus on skills that can’t easily be replaced by AI, such as communication and collaboration.”
A government spokesperson acknowledged the growing influence of AI, stating: “Generative AI has great potential to transform education… but integrating AI into teaching, learning and assessment will require careful consideration.”
