AI could replace 3m low-skilled jobs in the UK by 2035

Published: 25 November 2025. The English Chronicle Desk. The English Chronicle Online

Up to three million low-skilled jobs in the UK could disappear by 2035 as automation and artificial intelligence transform the labour market, according to a report published this week by the National Foundation for Educational Research (NFER). The research, based on long-term economic modelling, suggests that while AI and technological advances will create some new roles, they are likely to disproportionately affect lower-skilled occupations, leaving millions of workers facing uncertainty about their employment prospects.

The jobs most at risk, the report highlights, are those in trades, administrative roles, and machine-operating positions. These roles, often characterised by routine or repetitive tasks, are particularly vulnerable to automation. Conversely, highly skilled professions such as engineering, healthcare, education, and advanced technology roles are expected to grow in demand as AI adoption increases workloads and augments human capabilities.

Overall, the report predicts the UK economy will add 2.3 million jobs by 2035. However, these new roles will not be evenly distributed, with professional, managerial, and technical positions forming the majority, leaving many low-skilled workers with limited opportunities. The NFER research therefore suggests that while the headline numbers indicate job creation, the reality for many is likely to be more precarious.

These findings contrast sharply with other research that suggests AI will hit highly skilled roles harder than lower-skilled work. For example, a study published by King’s College in October reported that higher-paying firms experienced job losses of around 9.4% between 2021 and 2025, much of which occurred after the public release of ChatGPT in late 2022. This research indicated that professions such as software engineering, management consultancy, and legal work were highly exposed to AI disruption.

The UK government has similarly identified management consultants, psychologists, and legal professionals among occupations “most exposed to AI,” while roles such as sports players, roofers, and bricklayers are considered less vulnerable. Despite this, the NFER report warns that entry-level roles in administrative and manual occupations remain at serious risk, reflecting an uneven impact of technological change across sectors.

Recent corporate examples appear to underscore these dynamics. Last week, law firm Clifford Chance announced it would be laying off approximately 50 business services staff at its London offices, equating to roughly 10% of that department. The firm cited AI-driven changes in workflows as one factor behind the reduction. Similarly, PwC’s head publicly revised hiring plans for the 2021–2026 period, explaining that AI and automation had altered workforce requirements, reducing the number of new roles needed.

However, Jude Hillary, one of the authors of the NFER report, cautioned that these immediate layoffs may not reflect the longer-term impact of AI. He suggested that current job losses attributed to automation may instead be driven by broader economic uncertainty, rising national insurance contributions, and risk-averse employers. “There’s this general uncertainty about where things are going, how long it takes to improve,” he said. “There’s lots of talk about AI and automation without any real substance about it. Lots of employers are worried about it.”

Hillary added that many organisations may simply be “sitting tight,” delaying hiring or reassigning roles while they evaluate AI tools, rather than immediately displacing staff. He emphasised that the overall effect of AI on the UK workforce is likely to be complex. Some professional roles may expand due to AI augmentation, while many entry-level and low-skilled roles could diminish, creating significant challenges for workers who need to adapt to a changing labour market.

“The additional jobs that we’re getting in the labour market tend to be professional and associate professional roles,” Hillary explained. “Displaced workers, the one to three million that we talk about in our report, face significant barriers to get back into the labour market. These are not simple transitions. Reskilling will be difficult, and the support systems are not yet fully in place to make that feasible for everyone affected.”

Education and training will be crucial to mitigate these effects. Hillary pointed out that vocational programmes and adult reskilling initiatives will need to be scaled up dramatically to equip displaced low-skilled workers with the qualifications required for emerging professions. However, he warned that the pace of AI adoption may outstrip the speed at which workers can retrain, creating a potential mismatch between supply and demand in the labour market.

The report also highlighted geographic disparities in vulnerability. Workers in regions heavily reliant on manufacturing, logistics, and administrative sectors are likely to experience higher displacement rates, while urban areas with concentrations of knowledge-based industries may see more job growth. “The impact will not be uniform,” Hillary noted. “Local economies with fewer high-skilled opportunities will need targeted interventions to support displaced workers.”

Industry experts have stressed that AI should be seen as a tool for augmentation rather than pure replacement. Some organisations are already using AI to streamline repetitive tasks, allowing employees to focus on higher-value work, but these benefits may primarily accrue to companies and highly skilled staff. Without coordinated government policies, there is a risk that low-skilled workers could be left behind in the transition to an AI-driven economy.

Labour unions have also expressed concern. The Trades Union Congress (TUC) warned that AI and automation could exacerbate existing inequalities unless accompanied by strong retraining initiatives, fair pay, and social protections for workers in vulnerable roles. The union highlighted that without proactive intervention, the gap between high- and low-skilled workers could widen significantly over the next decade.

Despite these risks, some economists argue that AI could ultimately create new types of jobs that do not yet exist. Innovations in green energy, healthcare technology, AI ethics compliance, and robotics may open opportunities for both high- and medium-skilled workers, provided that proper training and education pathways are available. However, the uncertainty around job creation timelines makes planning challenging for policy makers, employers, and workers alike.

The NFER report concludes that while AI is likely to reshape the UK labour market dramatically by 2035, the narrative of mass, indiscriminate job destruction may be overstated. Instead, it recommends careful planning, investment in reskilling programmes, and policies aimed at equitable distribution of opportunities to ensure that the benefits of AI are shared across society.

“The conversation about AI-driven job loss often focuses on fear and disruption,” Hillary said. “But if we understand the nuances – which roles are likely to grow, which are likely to shrink, and how workers can be supported – we have the chance to shape a more resilient and adaptable workforce. It won’t happen automatically, but with foresight, the labour market can evolve alongside technology rather than be swept aside by it.”