Over Christmas, I was chatting to a relative who works in finance. Their company, a relatively small firm, had just opened applications for its summer internship in Manchester. “We had 600 applications in less than three days, for a single internship spot”, they told me. The scale was unprecedented and, most bizarrely of all, at least 200 of those candidates had applied from the same place: Texas, in the US – a hallmark of AI application en-masse.
Reports of the harms of AI in graduate recruitment are not unusual, either. According to The Times, UK entry-level jobs have reduced by 45% since the release of ChatGPT, and every day, thousands of bots are scraping websites and submitting applications indiscriminately. As my mind turned towards my own imminent graduation, I was disheartened to hear that even smaller companies aren’t safe.
It made for a less-than festive end to our conversation about graduate recruitment. “It’s a nightmare”, we agreed. But then, after a shrug, my relative added: “We might just have to go back to hiring people we know.”
I was taken aback by their candour. Nepotistic recruitment has never really gone away, especially in sectors like finance, but companies at least made an effort to pretend it had. Name-blind CV reviews and declarations of previous contacts at the company are interventions meant to level the playing field. To now return to an ‘oldboys’ recruitment method seemed like a huge step back.
But there’s a bleak kind of logic to it. You, as an employer, have no way of telling which applications are genuine human effort and which aren’t, and you don’t have the time or human resources to even read them all anyway. Your corporation, like most, doesn’t have an entire team dedicated to recruitment. It’s wasted work for you and the grads that actually did put the work in and whose applications won’t be given the attention they deserve.
So next year, instead of advertising the opportunity widely, you hire your next door neighbour’s son, who you’ve known since he was six and is a stand-up lad. Or you take on a grad from your alma mater, a rigorous university which you know produces hard-workers. With any luck, your neighbour will return the favour by offering your very capable niece an internship. And so it continues. Decades of progress in widening access reversed.
The only alternative appears to be a total embrace of AI on the other side of the hiring process. The vast quantity of AI-generated applications require the processing-power only AI can provide. Recruitment becomes reduced to chatbots applying to chatbots: dead internet theory at its finest.
And on top of that, companies specify that applicants shouldn’t use AI in their submissions, but why would they listen? AI will almost definitely be used to judge them, and given the number of job applications needed to secure a single graduate spot, students are choosing quantity over quality. Who can blame them?
Of course, there are options besides nepotism and AI armageddon. Hiring teams could specify criteria to reduce the overall number of applications and boost those from underrepresented groups: state-school educated, BAME, those with disabilities. Such “positive action” schemes have already been implemented successfully at a number of big corporations, including NatWest, The Guardian, and PwC. Though it doesn’t solve the issue for students not included within those groups, maybe it’s a start.
But do we really expect that most UK companies will make the effort? Their bottom line is their bottom line: they don’t really care about access. Grads without connections are collateral: unlucky – try again next year. I tentatively suggested the positive action policy to my relative, who seemed to just sigh hopelessly: “AI’s just wrecking the whole thing anyway.”
Which leads me to my final point, we often hear people using AI as a catch-all bad guy for the UK’s problems: “It’s all AI’s fault.” But AI itself can’t be at fault, AI doesn’t have agency: it doesn’t have consciousness (yet). For the time being, at least, humans are still behind the bots – there are people pulling the strings and making hiring policies. Just resigning to the idea that AI is taking over won’t solve anything. It’s more than depressing: it’s disempowering, and this kind of fatalism can only result in lazy policies.
In the meantime where does that leave the grads? Is the answer really just… network more? Make nepotism work for you? It’s hardly a fair or realistic conclusion. But if that is where recruitment is heading back to, there doesn’t seem to be much other choice. Oxford is privileged in the sense that its alumni networks are amongst the best in the world; many of the big societies, such as LawSoc, Oxford Women in Business, or the Union, have direct lines to some of the best in the business. I feel uneasy suggesting such a solution, especially given the barriers to entry to those societies in the first place, and the fact that Oxford students still struggle to get grad spots in spite of the promised leg up by the university brand. To say we can do absolutely nothing to help ourselves, though, feels like giving up.
But the real change can only come from employers, who must be emboldened to take meaningful action. It’s not over yet, and a real solution would benefit everyone involved. Because, lest we forget, there are still humans behind the chatbots applying to chatbots: students who just want prospects. Nepotism doesn’t provide that, aside from for the lucky few, and AI only complicates an already-crowded space.
Students want to put in the effort. We just need a fair chance – only human choices, not AI, can give that to us.
 they know. Is this a new age of nepotism?){kind=link}