AI has experienced massive growth in the past few years, and its rise has disrupted not only global markets and jobs but also society as a whole.
Some experts posit that AI will steal 99% of jobs by 2030 — don’t let that fool you into thinking that company profits will do anything but soar — while a recent report from Stanford University found that AI is mostly stealing jobs from young professionals.
AI is also affecting privacy and safety. In one extreme case, which has resulted in a lawsuit against ChatGPT‘s creator OpenAI, 16-year-old Adam Raine is alleged to have committed suicide after “months of encouragement from ChatGPT.”
While AI is affecting everyone, there are specific concerns surrounding Gen Z and education. Now, a new report from YouGov.uk lays out how UK students are actually using AI.
The study asked more than 1,000 university and undergraduate students a number of questions related to AI, and the results point, perhaps unsurprisingly for Gen Z, to rapid adoption.
While the data suggests that a majority of students are using AI in what I’d call a responsible manner, there are some standout numbers related to cheating, hallucinations, and overall preparedness from educational institutions.
Two-thirds of polled UK students use AI for study
Of the students polled, YouGov found that 66% use AI for study-related purposes; 33% of all students say they use AI at least once a week.
That leaves roughly a third of students who don’t touch AI at all for studies; beyond that, about 23% say they don’t use AI at all, for studies or otherwise.
ChatGPT is by far the most common AI model used by students, eating up 74% of the 691 students who said they use AI for study. Google’s Gemini comes in at second place with 11%, and Microsoft’s Copilot falls to third place at 8%.
AI can be useful in many different ways, but 81% of students who use it for studying say it’s most commonly employed to better explain difficult concepts. The second most common use, at 69%, is to summarize content for easier reading.
Leaning on AI to help with coursework certainly seems to be helping the UK’s students. Roughly 30% believe that AI has helped them achieve higher marks, while 11% believe that their marks have actually dropped since AI became so common.
Most students polled (36%) say that AI helped them learn and develop “somewhat more,” with 8% pushing that to “much more.” In terms of post-grad prospects, 12% of all students polled believe that AI tools will be “very important” in future jobs.
AI is contributing to cheating in UK universities
With AI’s use comes newfound concerns of cheating. YouGov’s study claims that 23% of students who say they use AI for study used it in at least one of three ways that could be seen as “potentially cheating behavior.”
These methods included using AI to create sections of coursework (20%), using AI to entirely create coursework before self-editing (12%), and using AI to entirely create coursework with no self-editing involved (5%). This coursework, of course, counted toward a student’s final grade.
It’s interesting to note that the 5% of students who say they fabricate coursework using AI and hand it in with no manual editing directly matches the percentage of students who believe using AI in this way is acceptable.
The 20% of students who use AI to create sections of graded coursework is nearly matched by the 19% of students who say this is acceptable. On the other hand, 78% of students who use AI claim that using it to improve graded coursework is acceptable, versus the 52% who actually follow through.
Two-thirds of all students polled believe that their university would “likely” detect AI cheating; only 24% believe it to be “very likely.” Out of all students, 24% believe that their university’s rules against the use of AI are “not strict enough,” with a majority 55% believing that the rules are “about right.”
Only 11% of all students say their university “actively encourages the use of AI in ethical ways, and has taught us skills to use AI well and how to avoid its pitfalls.” Another 15% claim that their university offers “little guidance,” while 18% say their university actively discourages the use of AI.
I wonder what those same students would have to say about the claims that British politicians are using ChatGPT to write speeches?
AI hallucinations remain a common occurrence
It’s no secret that AI isn’t always right. The horrific story I mentioned earlier involving Adam Raine is a prime example of AI delusion, but it’s certainly not alone.
OpenAI CEO Sam Altman has even warned that ChatGPT “should be the tech that you don’t trust THAT much,” a statement that flies in the face of much of the hype Altman has spewed over the years.
Getting misinformation from AI while casually using the tool is one thing, but it’s another thing entirely when viewed from an education angle. In this case, young students are more aware of these hallucinations than the general public.
According to the YouGov survey, 47% of university students who say they use AI for study claim that they often notice AI hallucinations. A separate YouGov poll called “How many Britons know AI lies?” suggests that 23% of general UK AI users have noticed hallucinations.
