September is the beginning of many young people’s lives, as cars speed along motorways transporting 18- and 19-year-olds to their new university accommodations. I remember my own journey down to Exeter in 2022, the first stage in what I hoped would be an experience to set me up for the rest of my life. Little did I know that this was the calm before the storm, before anyone had heard of ChatGPT, or imagined the chaos that generative AI was about to cause for new graduates.
Fast forward to 2025, and some of the young people I began this journey with have realised that they’ve spent the last three years training for graduate jobs that don’t exist. Many firms are now slashing their number of new hires. Big accountancy firms have cut back on graduate recruitment; Deloitte reduced its scheme by 18%, while EY has cut the number of graduates it’s recruiting by 11%. According to data collected by the job search site Adzuna, entry-level job opportunities in finance have dropped by 50.8%, and those for IT services have seen a decrease of 54.8%.
The main cause of this is artificial intelligence, which is destroying many of the entry-level jobs open to recent graduates. Companies are now relying on AI to replicate junior-level tasks, removing the need for them to hire humans. It feels like a kick in the teeth to students and recent graduates, who were already entering a challenging labour market. Once, graduates who had toiled through multiple rounds of interviews, battled it out with other applicants at an assessment centre, and made it through to the final round, could hope to get a job in a sector such as consultancy or accountancy. These historically secure, solid and (some would say) boring options guaranteed you gainful and well-paid employment and a clear career path.
Now, those secure opportunities feel as though they’re evaporating. Since applicants can’t see jobs that no longer exist, their experience of this intense competition for fewer jobs is often limited to a series of disappointments and rejections. Should a student or recent graduate apply for one of these elusive opportunities, their application will frequently be evaluated and often declined by an AI system before a human even reads it. Friends who have recently graduated tell me of the emotional toll of talking to their webcam during an AI-generated interview in the hope that the system judges in their favour, a process that can be repeated again and again.
So far, creative fields, and those that involve real-life human contact, seem more impervious to this trend. It will probably be a period of time before doctors or nurses, or professions that rely on genuine creativity such as painters or performing artists, find themselves replaced with an AI model. Even so, if people become increasingly unable to spot AI, and businesses continue to embrace it, the risk is that professions such as art and illustration also get devalued over time, and replaced by a bleak, AI-generated cocktail of eerily familiar “creative” work.
Conservative politicians and the rightwing press have often suggested that the most valuable degrees are those that have a clear job at the end of them (and that those in more creative fields, such as the humanities, are by implication less valuable). As one Times columnist wrote recently, students who do “less practical” degrees are more likely to be “living at home, working on their script/novel/music/art portfolio while earning pocket money”, without either a profession or a useful skill.
But what use is a degree in accountancy if you can’t then get an accounting job at the end of it? Why is this course more valuable than studying something that teaches you critical thinking and transferrable skills – anthropology, say, or (in my case) Arabic and Islamic studies? Cuts to higher education mean that we’re already seeing the end of some of those degrees often labelled as “useless”, yet the supposedly “useful” subjects start to look less valuable when the jobs associated with them are replaced by AI models that didn’t take three years to learn these skills.
The end of university is already a terrifying time. Three or four years of preparing a bulletproof LinkedIn profile and creating a plan for the future suddenly becomes real. The last thing a person needs aged 21 is for an AI model to take the job they were told their degree was essential for. Today the playing field that exists is different to that of a year ago, and it will undoubtedly be different again when I and many other students graduate in a year’s time. The adults who implore us to embrace AI to streamline everyday tasks and improve the efficiency of the working day often already have working days, a promise that feels as though it’s drifting further and further away.
