With the final Milburn Report on young people not in education, employment or training (NEET) due to be published imminently, Josh Hillman asks how AI could help young people find their way into education and work – and what risks it poses for those already facing disadvantage.
Preparing young people for an AI-driven future
Artificial intelligence (AI) is rapidly moving up the education and skills policy agenda. The UK government wants young people to develop the digital and AI capabilities they need for life and work. At the same time, policymakers, parents and educators are concerned about screen use, digital distraction, online harms and technology’s effects on learning and development.
These are not necessarily contradictory positions. But they demand a more sophisticated conversation than simply advocating more or less technology. The challenge is to understand how AI can support young people’s learning, development and employment prospects without reinforcing existing inequalities or replacing the human support they need.
This conversation is particularly urgent as government seeks to strengthen routes into employment and tackle youth inactivity. Skills England has highlighted AI adoption and young people’s employability among its major challenges, while a new Early Careers Jobs Alliance is examining how employers can preserve routes into work. Government is also exploring the role of generative AI in education, including the potential for AI tutoring tools to support disadvantaged pupils, while acknowledging that evidence on pupils’ use of generative AI remains embryonic.
Against this backdrop, the final Milburn Report will bring renewed attention to the scale and complexity of youth disengagement. The question is not simply what AI means for the future of work, but how young people who are already at risk of being left behind can be supported to benefit from the opportunities it creates.
Tackling disengagement early
The term ‘NEET’ describes a highly diverse group of young people with different levels of attainment, needs and barriers to participation. While the existing cohort is extraordinarily large, the longer-term priority must be prevention.
This means tackling disengagement early and providing sustained support through the early years, school attendance, special educational needs provision, mental health support, basic skills training, post-16 pathways, careers guidance and employer engagement.
The potential impact of AI on youth disengagement is difficult to separate from wider labour market and social trends. The Interim Report of the Milburn Review highlighted the barriers facing young people alongside a structural weakening of routes into work, including falling youth employment, fewer entry-level opportunities and increasingly demanding recruitment processes.
There is a further concern about the nature, not simply the number, of entry-level jobs. If AI automates routine tasks through which young people traditionally learned on the job, employers may find it harder to develop the experienced staff they later rely on.
As AI reshapes the labour market, there is also a risk that these structural challenges become even more pronounced for young people who are already struggling to access education, training and employment.
How AI could deepen existing inequalities
AI could make existing disadvantages more consequential in four ways.
First, young people disconnected from education and employment may have fewer opportunities to develop digital literacy and the skills needed for work. As AI becomes more embedded in recruitment, workplaces and education, this could further widen the gap between those who have developed these capabilities and those who have not.
Second, socio-economic disadvantage can limit access not just to devices and connectivity, but also to the skills, confidence and support needed to use technology effectively. Access to AI alone does not guarantee that young people will be able to use it to improve their learning or employment prospects.
Third, AI could deepen inequalities in learning and development. More advantaged young people may be better placed to use AI as a tool for learning and coaching. For those facing greater disadvantage, there is a risk that AI becomes a lower-cost substitute for human teaching, training or support. This raises important questions about the quality and equity of provision, particularly for young people who need more tailored support.
Finally, AI may intensify what the Milburn Review describes as ‘rejection at scale’. Entry-level recruitment has become increasingly remote and formalised, with online portals, automated screening and, in some cases, AI-mediated recorded interviews.
For young people furthest from education and employment, these processes can create additional barriers. Repeated rejection without human contact or meaningful feedback may further undermine confidence and self-esteem, making it harder to re-engage with the labour market.
The potential for AI to open up opportunities
Despite these risks, AI also offers opportunities to improve outcomes for young people who are NEET or at risk of disengagement.
Better use of data could enable earlier identification of young people at risk, allowing schools, colleges and support services to intervene before disengagement becomes entrenched. AI could also make information about education, training and employment opportunities more accessible, helping young people navigate what can be a complex landscape of pathways and support.
Research by the Institute of Student Employers involving young people outside education and employment identifies potential uses of AI to help navigate recruitment processes and communicate transferable skills. These applications could help young people better understand their strengths, identify relevant opportunities and overcome some of the practical barriers to entering the labour market.
But information is not the same as guidance. AI may make options easier to find, but young people still need trusted human support to make sense of them, recognise their strengths and connect to real opportunities.
Apprenticeships, T-Levels and V-Levels could become particularly important where they combine formal learning, mentoring and workplace experience, with employers involved in their design. Such approaches could help ensure that young people develop the skills employers need while retaining access to meaningful, supported routes into work.
However, the education system cannot predict precisely what the labour market will need in the years ahead. Preparing young people for an uncertain future will require strong foundational knowledge, durable employment skills, genuine AI literacy, high-quality careers guidance and access to tools that are appropriate, ethical and effective.
Assessment will also need to reflect and reinforce these capabilities, ensuring young people develop the knowledge and judgement required to use AI well rather than becoming dependent on it. Used well, AI can scaffold learning; used badly, it can allow young people to bypass some of the reasoning and effort through which knowledge and judgement develop.
Putting young people at the centre
The challenge for policymakers, educators and employers is to ensure that AI supports rather than replaces the human relationships, tailored support and meaningful opportunities young people need to thrive.
This means looking beyond technology itself to the wider conditions that enable young people to participate in education and employment. AI could help identify those at risk, support learning and open up new pathways into work. But its benefits will not be automatic.
The priority must be to ensure that young people who face the greatest barriers are not further disadvantaged by the technologies reshaping their futures. Understanding how AI can be used fairly and effectively will be an important part of building better pathways into education, training and employment.
By Josh Hillman, Director of Education, The Nuffield Foundation