Is AI behind youth unemployment?

“According to the ONS, youth unemployment has jumped to uncomfortable levels, higher than the EU average. Grace Lordan untangles the many causes behind the rise, including AI and a rising minimum wage, and offers proposals for government action. Enjoying this post? Then sign up to our newsletter and receive a weekly roundup of all our articles. The latest Office for National Statistics (ONS) figures make for uncomfortable reading. Youth unemployment has risen to 16 per cent , with 732,000 people aged 16 to 24 out of work, nearly 100,000 more than a year ago. Close to a million young people are now classified as NEET (not in education, employment or training). UK youth unemployment has now overtaken the EU average for the first time since comparable records began in 2005. So what is driving this? The honest answer is that there is no single villain. But there are three forces converging in a way that should alarm anyone who cares about social mobility and long-term economic health. The AI effect is real, but not in the way you might think I remain broadly optimistic about what AI will do for work in the long run. Technology has historically complemented human labour more than it has displaced it. My research with Cecily Josten found that jobs requiring “people” engagement combined with abstract thinking are the safest from automation, and that the skills powering these roles are fundamentally human: social perceptiveness, negotiation, persuasion, and care. But the short-run adjustment period we are living through is genuinely painful with winners and losers. The losers are disproportionately young. The tasks that are being automated first are precisely the tasks that used to constitute a young person’s first job. Entry-level vacancies across the UK have dropped by roughly a third since late 2022. A timeline that maps uncomfortably neatly onto the arrival of generative AI tools in the mainstream. A recent British Chambers of Commerce study found that 54 per cent of small and medium-sized enterprises (SMEs) are now using AI, more than double the figure from two years ago. A British Standards Institution survey of more than 850 business leaders found that 43 per cent expect to reduce junior roles or entry-level hiring in the coming year due to AI-driven efficiencies. The tasks that are being automated first are precisely the tasks that used to constitute a young person’s first job: drafting, data entry, basic research, scheduling, customer queries. These were never glamorous. But they were the way people got their foot in the door, learned how organisations work, and began building the professional judgement that no qualification can teach you. This is the real danger. The issue is not just that fewer entry-level jobs exist today. It is that we risk dismantling the bottom rungs of the career ladder entirely. In my work with David Neumark , I showed how minimum wage increases interacted with automation to reallocate workers out of automatable jobs. The pattern we documented then is accelerating now, but with a new technological driver. The cost of hiring the young has gone up The second force is the rising cost of employing young workers. The National Minimum Wage for 18-to-20-year-olds jumped by a record 16.3 per cent in April 2025, from £8.60 to £10.00 and has just risen again to £10.85 (a further 8.5 per cent increase). The government has been explicit that it intends, over time, to align the youth rate with the adult National Living Wage. From a fairness perspective, this is entirely defensible. Young people face the same cost of living as everyone else. They should not be penalised by law for being twenty rather than twenty-one. But we have to be honest about the trade-offs. Research by the National Institute of Economic and Social Research suggests that recent policy changes (including National Insurance contributions (NICs) increases, minimum wage rises, and reforms to employment rights) have raised the real cost of hiring entry-level workers by around 7 per cent. For young people who have never managed to get a foothold in the first place, the psychological toll of perpetual rejection in a shrinking market should not be underestimated. When you combine a more expensive young worker with an AI tool that can do a meaningful share of their tasks for a fraction of the cost, the economic logic for some employers becomes brutally straightforward. The Resolution Foundation has warned that extending the top wage rate to 18-year-olds risks making the youth employment situation go from “bad” to “worse.” This is not an argument against fair pay. It is an argument for recognising that wage policy does not operate in a vacuum and employers will react to rising costs. The sickness crisis is hiding in plain sight There is a third, less discussed factor. According to House of Commons Library analysis , the number of 16-to-24-year-olds who are economically inactive due to long-term sickness has nearly doubled in a decade, rising from around 138,000 to 271,000. The Health Foundation reports that the share of NEET young people reporting a work-limiting health condition rose from 26 per cent to 44 per cent between 2015 and 2025. This is a 70 per cent increase in a decade. Many young people are not unemployed in the traditional sense. They have withdrawn from the labour market altogether, often because of mental health conditions that worsened during and after Covid. My research with Eliza-Jane Stringer on automatable work and mental health found that even being in a job at high risk of automation has a detrimental effect on workers’ mental health and life satisfaction. The anxiety produced by precarious, uncertain work is not confined to those who have already lost their jobs. It radiates outward. For young people who have never managed to get a foothold in the first place, the psychological toll of perpetual rejection in a shrinking market should not be underestimated. This matters for policy design. A hiring subsidy will not help a 19-year-old who is too anxious to leave the house. The government’s recently announced £1 billion package , including a £3,000 Youth Jobs Grant for employers and an expansion of the Youth Guarantee to cover 18-to-24-year-olds, is a welcome step. But as the Institute for Fiscal Studies has argued , in practice these policies are likely to directly benefit only a small fraction of the almost one million 16-to-24-year-olds who are NEET, with around 320,000 eligible young people on Universal Credit not required to search for work, mostly due to health reasons. If we design interventions only for the job-ready, we will leave the most vulnerable behind. What should the government do? First, invest seriously in the transition phase of the fourth industrial revolution, not just the destination. The current policy debate is too binary: either AI will create wonderful new jobs, or it will destroy everything. The reality is that the adjustment period between old jobs disappearing and new ones emerging is where the damage gets done. Government needs to fund retraining infrastructure that moves at the speed of technological change, not at the speed of Whitehall. The new AI and automation practitioner apprenticeship is a start, but we need much more. This includes digital skills embedded in every employment support programme, not bolted on as an afterthought. Second, redesign entry-level work rather than eliminating it. Firms should be thinking about how AI can augment a junior employee rather than replace them. A graduate working alongside AI tools can learn faster and contribute more quickly, but only if the role is designed with development in mind. Government can nudge this through the tax system and through procurement standards that reward firms maintaining early-career pipelines. Third, take the mental health crisis among the young as seriously as the employment crisis, because they are the same crisis. Integrated support that addresses health and employability together, rather than parking people on benefits or funnelling them into programmes they are not ready for, is essential. The youth work sector has good evidence on what works here, and it deserves sustained investment rather than short-term project funding. Fourth, do not lose sight of the skills that matter. As I have argued in work on the accelerated value of social skills and in a broader study on the changing demand and return to skills for professional workers , the capabilities that are hardest to automate (social intelligence, adaptability, creative problem-solving) are precisely the ones our education system does least to cultivate. If we want young people to thrive in an AI-augmented labour market, we need to invest in these distinctly human skills from school onwards, not treat them as a nice-to-have. Last, do not lose sight of inclusion. The pain of this labour market is not evenly distributed. Young people from lower socioeconomic backgrounds, ethnic minorities, and those outside London face compounding disadvantages. A graduate from a Russell Group university with parental support and professional networks will eventually find work. A school leaver in a post-industrial town with no connections and deteriorating mental health may not. Policy must be designed for the latter, not the former. The UK has been here before. Youth unemployment was higher in the early 2010s and came down. But the structural conditions are different now. AI is not a recession. It does not pass. The entry-level jobs it is absorbing are unlikely to come back in the same form. If we do not act with both urgency and intelligence, we will not just have a lost generation. We will have permanently reshaped who gets to build a career in this country, and who does not. Enjoyed this post? Sign up to our newsletter and receive a weekly roundup of all our articles. All articles posted on this blog give the views of the author(s), and not the position of LSE British Politics and Policy, nor of the London School of Economics and Political Science. Image credit: on Shutterstock The post Is AI behind youth unemployment? first appeared on LSE British Politics .

Summary generated from the RSS feed of LSE British Politics and Policy. All article rights belong to the original publisher. Click through to read the full piece on blogs.lse.ac.uk/politicsandpolicy.