95% of UK undergraduates use AI in at least one way, according to the Higher Education Policy Institute.
Are you one of them? Maybe you use it to summarise academic papers that make you yawn. Maybe you’ve asked ChatGPT to explain a tricky concept to you like you’re 5. Maybe you’re part of that 5% that doesn’t use AI at all.
Regardless, before the age of AI, you were probably doing fine academically. More than fine, even: you thrived. You thrived because you knew how to study, to work hard, to persist. It’s what got you to where you are now. So, why short-circuit a process you’ve spent years mastering?
A tutor that spits out solutions in record time, is at your disposal 24/7, and does it all free of charge sounds almost too good to be true. Since a human couldn’t possibly fit the bill without breaching tens of labour laws, for many UCL students, AI is the one for the job.
I spoke to several coursemates to better understand how they used AI.
Some described using it to gain tailored feedback on their coursework before submission. “It’s good to have a second view on things”, says Alma, a Biological Sciences student. Others use AI to simplify difficult concepts and generate more practice questions.
These conversations made it clear to me that students aren’t using AI to get out of doing work. They’re using AI to get more out of the work they already do. It offers a way to make the learning process more efficient - and less of a headache.
But a headache may not be the only thing that AI chatbots help us avoid. Research by Dr Colin Fu, one of UCL’s own educators, and Dr Athina Ioannou offers one possible explanation for why many students keep coming back to AI: “Techno-Insecurity”. When students use AI to generate ideas, essays and solutions, its polished outputs can make their own raw work seem inferior by comparison.
By continuing to use AI, students are shielded from the shame of “human failure”: of staring at a blank document and not knowing how to begin, of getting a question wrong and having to sit with that. Are we turning to AI because we’ve become less willing to be bad at something before we can be good at it?
It’s an easy trap to fall into. Maybe you’re in a lecture and you’ve lost track of what your lecturer is saying. You look to the people around you, hoping to see the same panic you feel inside reflected on their faces. To your left, one friend is furiously typing away. To your right, another is completely absorbed in the lecture.
Great.
The lecture ends, and you go to the library with your friends. You know you have questions. You also know that you’d rather not expose yourself as the one who doesn’t ‘get it’. So, you open your laptop, upload your lecture slides, and ask ChatGPT. You get answers in seconds. That niggling feeling in your gut of not knowing goes away before you have to sit with it for too long. The threat is eliminated. You’re safe again for now.
Perhaps that discomfort feels so intolerable because we’re under more pressure than ever to succeed.
With youth unemployment rates being at their highest in over a decade, and CVs now being screened by bots rather than real people, the future looks bleak for Gen Z students. And it seems like there is less and less room to make mistakes; to achieve anything less than a 1st (or God forbid, a 2:1), to not get that internship, to not have a perfect LinkedIn profile, to not have enough extracurriculars on your CV. Less room to be human.
The state of our economy makes every setback we face feel like a direct threat.
For those without a safety net to fall back on, the pressure is even greater: education can feel like the only route to financial security and independence. For many, university has become something to get through – a series of boxes to tick – rather than an experience to immerse themselves in. And when box-ticking takes priority, AI becomes your best friend. This rise in AI usage amongst students could simply be a response to the uncertainty we feel about our futures; some students may cling to AI, hoping that it will give them certainty in these tough times.
For some students, the things AI helps us avoid are precisely what make education worthwhile. “I can very confidently say I don’t need AI," confessed Peggy, a Biological Sciences student.
“I think a necessary part of education, and university in general, is the commitment of time and effort to work”.
She raises an important question: what’s at stake when we remove the very struggle that education requires?
While AI’s novelty means that the dust has yet to settle on its long-term effects, researchers have some ideas, and they aren’t very encouraging. An early but striking 2025 study from MIT Media Lab - based on a small sample of 54 participants - found that students who used LLMs while writing essays showed weaker neural connectivity than those who wrote without them. Despite the study’s preliminary nature, its findings hint that outsourcing too much of the learning process may come at a cost.
The more we delegate the difficult parts of learning to chatbots, the less practice we get at solving problems, questioning information and thinking for ourselves. Ironically, using AI to become the cookie-cutter graduates we think employers want may instead erode the very skills that make us employable. AI will only get smarter, faster and better over time.
The question is whether we’ll become better alongside it - or simply more dependent.
With academic deadlines stacking up, and the threat of postgraduate unemployment looming in the back of our minds, it’s hardly surprising that even the students who need AI’s help the least are the ones reaching for it. However, if you do choose to use AI in your studies, just make sure that you know where the bot ends and you begin. Make sure that you can still pull the plug.