Revealed: Thousands of UK university students caught cheating using AI
Thousands of university students in the UK have been caught misusing ChatGPT and other artificial intelligence tools in recent years, while traditional forms of plagiarism show a marked decline, a Guardian investigation can reveal.
A survey of academic integrity violations found almost 7,000 proven cases of cheating using AI tools in 2023-24, equivalent to 5.1 for every 1,000 students. That was up from 1.6 cases per 1,000 in 2022-23.
Figures up to May suggest that number will increase again this year to about 7.5 proven cases per 1,000 students – but recorded cases represent only the tip of the iceberg, according to experts.
The data highlights a rapidly evolving challenge for universities: trying to adapt assessment methods to the advent of technologies such as ChatGPT and other AI-powered writing tools.
In 2019-20, before the widespread availability of generative AI, plagiarism accounted for nearly two-thirds of all academic misconduct. During the pandemic, plagiarism intensified as many assessments moved online. But as AI tools have become more sophisticated and accessible, the nature of cheating has changed.
The survey found that confirmed cases of traditional plagiarism fell from 19 per 1,000 students to 15.2 in 2023-24 and is expected to fall again to about 8.5 per 1,000, according to early figures from this academic year.
The Guardian contacted 155 universities under the Freedom of Information Act requesting figures for proven cases of academic misconduct, plagiarism and AI misconduct in the last five years. Of these, 131 provided some data – though not every university had records for each year or category of misconduct.
More than 27% of responding universities did not yet record AI misuse as a separate category of misconduct in 2023-24, suggesting the sector is still getting to grips with the issue.
Many more cases of AI cheating may be going undetected. A survey by the Higher Education Policy Institute in February found 88% of students used AI for assessments. Last year, researchers at the University of Reading tested their own assessment systems and were able to submit AI-generated work without being detected 94% of the time.
Dr Peter Scarfe, an associate professor of psychology at the University of Reading and co-author of that study, said there had always been ways to cheat but that the education sector would have to adapt to AI, which posed a fundamentally different problem.
He said: '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.'
Students who wish to cheat undetected using generative AI have plenty of online material to draw from: the Guardian found dozens of videos on TikTok advertising AI paraphrasing and essay writing tools to students. These tools help students bypass common university AI detectors by 'humanising' text generated by ChatGPT.
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.'
Harvey* has just finished his final year of a business management degree at a northern English university. He told the Guardian he had used AI to generate ideas and structure for assignments and to suggest references, and that most people he knows used the tool to some extent.
'ChatGPT kind of came along when I first joined uni, and so it's always been present for me,' he said. 'I don't think many people use AI and then would then copy it word for word, I think it's more just generally to help brainstorm and create ideas. Anything that I would take from it, I would then rework completely in my own ways.
'I do know one person that has used it and then used other methods of AI where you can change it and humanise it so that it writes AI content in a way that sounds like it's come from a human.'
Amelia* has just finished her first year of a music business degree at a university in the south-west. She said she had also used AI for summarising and brainstorming, but that the tools had been most useful for people with learning difficulties. 'One of my friends uses it, not to write any of her essays for her or research anything, but to put in her own points and structure them. She has dyslexia – she said she really benefits from it.'
The science and technology secretary, Peter Kyle, told the Guardian recently that AI should be deployed to 'level up' opportunities for dyslexic children.
Technology companies appear to be targeting students as a key demographic for AI tools. Google offers university students a free upgrade of its Gemini tool for 15 months, and OpenAI offers discounts to college students in the US and Canada.
Lancaster said: '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.'
A government spokesperson said it was investing more than £187m in national skills programmes and had published guidance on the use of AI in 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.'
*Names have been changed.
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Times
an hour ago
- Times
I have motor neurone disease — and the fund to find a breakthrough
Tris Dyson knew something was wrong in June 2022 when he could not move his left thumb. Six months later, aged just 44 and with a baby daughter, he was diagnosed with motor neurone disease (MND). The muscles in his body were dying and he would, sooner or later, lose the ability to walk, speak and breathe. 'I was newly engaged. I'd bought a house. Things were going quite well. So to be told the party's over was not brilliant,' recalls Dyson. Today he considers himself 'one of the lucky unlucky ones' as his physical decline is 'relatively slow'. Having being told he had four years to live, Dyson said he is in 'pretty good shape', and thinks the doctor's prediction was wrong. Since his diagnosis, Dyson, now 46, has focused his energies on his day job, which just happens to be organising multimillion-pound prizes for scientific breakthroughs. He is managing director of the Longitude Prize, which was established in 2014 as modern-day version of the original Longitude Prize of 1714. The original prize sought reliable ways of measuring longitude at sea, and it took nearly 50 years to be solved. The modern-day version has run two awards so far: the first, in 2014, to tackle antimicrobial resistance, and the second, launched in 2022, seeks treatments for Alzheimer's. Now, the next prize will target MND; specifically amyotrophic lateral sclerosis (ALS), the most common type, which is the one that Dyson suffers from. On Wednesday, Dyson's organisation, Challenge Works, a subsidiary of Nesta, the UK scientific foundation, will launch a £7.5 million Longitude Prize to find a treatment for ALS. For the next five months, Dyson will seek applications from medical researchers and AI experts all over the world before drawing up a shortlist of 20 teams. These will be whittled down in stages, with tranches of money doled out at each step. 'Prizes are very good when you've got a problem and you don't know where the solutions are going to come from,' Dyson says. 'You award on the basis of success.' The winning team will receive £1 million at the end of the challenge. Dyson hopes other investors or private companies will support the prize over time, offering computational power, for example, to run AI calculations. After that, the aim would be for a big pharmaceutical company to take the research and develop a drug: 'We need to feed them with high-potential, credible targets.' Pharma companies would have 'a huge market, potentially', he says. MND is not as rare as you may think. 'About one in 300 people will get it,' Dyson says. 'It's absolutely extraordinary. But often it appears when somebody is very old and it's part of the end-of-life process.' The Motor Neurone Disease Association estimates that it affects up to 5,000 adults in the UK at any one time. • MND sufferer records voice bank so she can swear at her husband MND is characterised by a breakdown in communication between nerve cells called motor neurones and the body's muscles, with the muscles weakening and wasting away, leading to paralysis. The progress of the disease varies from person to person, with only very limited treatments available and no cure. It is more likely to affect people aged over 50. There is no clear cause for MND, although genetic and environmental factors are thought to play a part. About one in ten people may have a family history which makes them more prone to the disease. Dyson believes there now is an opportune moment to tackle MND because of leaps in our understanding of genomics, and the rise of AI, which can sift through big data to find targets for drugs. Until now, he says, research has been limited and diagnosis is still pretty much the same as it was 100 years ago. 'What's happened in the last 15 years is the basic level of science has gone from being very little to a strong basic level of understanding,' Dyson says. 'The other thing that's happened is a huge amount of patient data has been collected, which we're making available through this prize.' Most of the money that has gone into studying MND has come from donations of patients' families, he adds, and cases such as that of the rugby league player Rob Burrow have raised its profile. Burrow and his former team-mate helped raise £6.8 million for a MND centre, which will open later this year in Leeds. 'It takes your independence away,' explains Dyson. 'It takes your dignity away. Some people manage to rise above it. Rob Burrow is a good example. It's still an unbelievably horrendous thing. And then you ask what treatments are available. There aren't any, which is an unbelievable thing to be told in this day and age.' Solving MND is not where Dyson expected his career to take him. He is not a scientist or medic by training; he studied Geography at UCL, where I first encountered him as an undergraduate, hanging around the ground floor of Ramsay Hall, a hall of residence in Fitzrovia whose main claim to fame is that the band Coldplay met there. After we graduated, I would see Dyson at various parties over the years. One early initiative was running a local currency scheme in Wales; another involved setting up a 'time credits' scheme across the UK that offered rewards to people who volunteered within a community. 'There was no plan,' Dyson says, looking back on his career, which brought him to Nesta in 2012. He has not, thankfully, lost his sense of humour, and talks calmly about the challenges ahead. 'I've got weakness in my arms. I might struggle to pick that up,' he says, reaching out to lift a jug of water —which he does, successfully. 'I'm OK walking. I have some problems typing.' He has not yet moved out of his home in Greenwich, southeast London, which has four flights of stairs, because he does not want to 'inconvenience' his partner, Jenny, and their daughter, now three. 'I'm putting money aside for my family. More than I would otherwise.' The unspoken question, of course, is whether Dyson will live to see the fruits of the prize he has founded. How does he look to the future? 'I'm not thinking too much about that,' he says. 'I think Stephen Hawking was like this. He just ignored it.' Hawking was diagnosed at 21 and died at the age of 76. What does he want people to know about MND, and his prize fund? Dyson pauses. 'We have to bring the timeframe of a treatment forward with this prize. I want people to believe it's solvable, and I think that's true.'
Daily Mail
an hour ago
- Daily Mail
When prompted to show a female investor, here's what AI created!
Artificial intelligence (AI) may now be able to do everything from creating images to providing answers to almost any question – but there is one area where it is stumped. It struggles to imagine women as investors. When asked to create images of investors, AI tools overwhelmingly depicted men. Even when asked to portray an investor with traditionally feminine characteristics – such as a skirt or painted nails – they still showed men, just with those features. That's according to trading and investment company eToro, which put AI platforms to the test. This is despite there being 6.7 million female investors in the UK – fewer than the 10 million male investors, but still a sizeable number. eToro asked so-called 'generative AI' tools to create the images. The platforms use algorithms – a set of instructions designed to solve a problem – to create content such as text or pictures. A user can type in a prompt, which may be a question or a description of what you want the tool to give you. For example, you may ask the AI tool to write a shopping list or to produce an image of a dog. It then uses online sources – which could include photographs, books, news articles, journals and other internet material – to answer in a conversational tone or create an image based on what you requested. But as it uses existing materials, the biases of its sources can creep into its responses to any requests. When asked to produce an image of an 'investor in a skirt', three of the four images created by the AI tools were sharply dressed men in skirts. Only one was of a woman in a pencil skirt. When the AI tools were asked to produce an image of an 'investor with red fingernails', all four pictures produced were of men in suits wearing red nail polish. It was only when asked to produce a hyper-realistic 'portrait of an investor in a dress' did all four AI platforms finally show an image of a woman. Unfortunately, some of these images show the investors wearing revealing clothing, while another woman wears a dress made from banknotes. AI also assumes women are assistants to investors. When AI was promoted to show 'an investor with their assistant', it created images of mostly middle-aged men in suits as the investor and women as their assistants. Lale Akoner, global markets Analyst at eToro, says: 'The misleading and harmful stereotype of the investor as a professional-looking man is sadly alive and well. 'The results simply tell women they don't belong. 'This isn't just an AI bias – it's a societal issue holding back women financially.' Depictions of investors in films, books and articles will be partly to blame as these sources are likely to be absorbed by AI tools. Dr Ylva Baeckstrom, a former banker and now senior finance lecturer at King's College London Business School, found male lead actors make up 76 per cent of the screen time allocated to lead roles in films about finance. The research, published by eToro, proves films mainly show men as the investor while women are depicted as wives, mistresses and assistants. It's a worry as households are starting to turn to AI for financial information. Dr Baeckstrom says women should question all of the information AI tools produce – and the first step in doing this is to become financially educated so you can analyse the information it gives you. Plus, this will help spot what is called 'AI hallucinations', where tools answer using false or nonsensical information. 'AI will make things up. You can't know whether it's true or not – you can't trust it. If you're a savvy user, you're much more likely to benefit from it. You'll question things,' she says. Finally, users must correct AI platforms if they produce a text or an image with these financial biases, Dr Baeckstrom says. 'We have the opportunity with AI to start again without the biases. We need to teach AI that the blueprint of an investor is not a man.'
Daily Mail
an hour ago
- Daily Mail
Ex-education secretary Michael Gove warns schoolchildren need more traditional classroom teaching to fend off dangers of AI 'cheating'
Former education secretary Michael Gove has said children don't need to be taught digital skills but need a proper academic education to help them combat the rise of AI. And he has argued that with 'an epidemic of cheating' triggered by AI, a handwritten essay is the surest test of ability. The Tory grandee, now Lord Gove, said he is more convinced than he ever was about the importance of a traditional education as AI tools like Chat GPT become commonplace. 'AI has created a huge potential for corruption and a huge opportunity for people to present themselves as knowledgeable when they don't actually have those skills. The way to demonstrate you have those skills is through that traditional knowledge. 'It is precisely because AI is going to change our lives so much that I believe we need to be armed with the knowledge that traditional education provides.' In his battle against what he saw as schools dumbing down, Lord Gove's tenure as Education Secretary in the coalition Cabinet from 2010 to 2014 was marked by his run-ins with education unions. Speaking to education leaders and pupils at an education conference at the Princess of Wales' former school Downe House in Newbury last week, he said children had got 'all the digital skills they could ever need'. And he told them it was 'better to live in the world of Pythagoras than Pokémon'. 'One of the in vogue buzz phrases of our era is "we want schools to teach digital skills". Nonsense! Young people today are digital natives – it's what they get outside school - arguably to excess.' Instead he said 'access to an academic education' would give children 'skills such as critical thinking, social confidence and the ability to understand what is true and false'. 'The future of education is going back to tradition. AI is already changing what is happening in education and we are already seeing its impact in higher education. 'An increasing amount of the work submitted is work that is not the product of scholarship or time spent in dusty libraries or at the thoughtful analysis of lecture notes but it's taken direct from chat GPT or other AI tools. 'There is an epidemic of cheating in higher education and the same thing could spread throughout the whole education ecosystem.' He said coursework could easily 'be downloaded or synthesised by AI' and delivering exam work through a digital platform was 'a retrograde step because those tools make it easier to cheat'. The former cabinet minister, now the editor of The Spectator magazine said the 'surest test in higher education and also in schools of what you can do and the ability to pass on to the next stage of life is actually a handwritten pencil and paper essay or an oral exam conversation'. He argued that young people needed 'the patient practice of skills and the cumulative acquisition of knowledge' to acquire the 'mastery' in a subject which would help them to use AI to their advantage. 'Who would have thought that this whole technological revolution would have reminded us that the traditional ways to teach are actually the best,' he said. Hitting out at 'the progressive educational argument' that 'because all that information is online or available online that you don't need to know as much as previous generations'. He said: 'The opposite is true – because there is so much information out there – you need to know more. You need more in order to be able to discern the bogus from the honest. 'A well-stocked mind is the best guide and the most effective guide for making your way through this forest of knowledge which is there for you.'
