Britain's employment problem will not be solved by training people for jobs that employers are increasingly reluctant to create. That is the central weakness in this Budget.
The Chancellor has recognised the importance of artificial intelligence, the scale of youth unemployment and the need to improve Britain's productivity. There are sensible measures in each area. Yet the policies remain disconnected from the decision that ultimately determines whether somebody gets a job: whether an employer believes taking them on is worth the cost and risk.
Having spent much of my career in recruitment and more recent years working with AI in employment, I believe that decision is becoming more difficult, particularly for businesses considering inexperienced candidates. The Budget offers some help, but it does not adequately confront the combined pressures facing employers or the possibility that AI will change which jobs they choose to create.
The Office for Budget Responsibility's latest forecasts provide an uncomfortable backdrop. Growth is expected to slow to 1 per cent in 2027, unemployment is forecast to average 5.1 per cent and business investment is expected to fall by 0.5 per cent. These are forecasts rather than certainties, and employment is expected to grow over the medium term. Nevertheless, they describe an economy in which businesses have limited confidence to expand.
That matters because recruitment is ultimately an investment decision.
An employer does not simply compare a candidate's salary with the work they might perform. They consider National Insurance, pensions, training, supervision, administration and the consequences if an appointment does not work out.
For a large employer, those costs can be absorbed across an established workforce. For a small business considering its fifth or tenth employee, they can determine whether a vacancy exists at all.
The Government's employment policies need to be understood in that context.
From April 2027, the National Living Wage will rise to £13.21 an hour. That increase will be welcome to workers struggling with household costs. Paying people properly is not incompatible with a competitive economy, and Conservatives should have no difficulty recognising the value of work.
But the statutory wage is only one component of the cost of employing somebody.
Employers are already operating with a higher rate of employer National Insurance and a lower secondary threshold than before April 2025, although the Employment Allowance provides important relief for eligible businesses. Workplace pensions and other obligations add further costs.
Employment regulation is also changing. The Employment Rights Act 2025 will reduce the qualifying period for ordinary unfair dismissal claims from two years to six months from January 2027 and remove the existing cap on compensatory awards. Other reforms are being introduced in stages, including changes affecting family leave and working arrangements.
These changes should be described accurately. Ordinary unfair dismissal protection is not becoming a universal day-one right. Nor should we assume that better employment rights necessarily destroy jobs. Clear rules and good management can improve retention, reduce disputes and benefit productive businesses.
The difficulty is that ministers tend to present each change according to its individual purpose. Employers experience the cumulative effect.
Consider the independent business owner facing higher staffing costs, rising supplier prices and expensive energy while trying to maintain affordable prices for customers. A reduction in business rates may be welcome, but it does not necessarily offset the pressures elsewhere in the business.
The Chancellor's promise of breathing space must be judged against the entire balance sheet, not one line of it.
A business deciding whether to recruit does not separate the wage bill from tax, compliance, training and legal exposure. It considers the whole commitment. If that commitment becomes less attractive, the response may not be an immediate redundancy. It may simply be a vacancy that is never advertised.
This is especially consequential for young people.
The first job is rarely the one in which somebody makes their greatest economic contribution. It is where they learn how a workplace functions, develop judgement, make mistakes under supervision and acquire skills that cannot be taught entirely in a classroom.
Employers accept an initial gap between a beginner's cost and productivity because they expect that investment to pay off over time.
If the economics of that investment deteriorate, inexperienced candidates are particularly exposed. An employer may seek somebody who can contribute immediately, delay recruitment or redistribute work among existing staff.
Artificial intelligence introduces another consideration into that calculation.
For years, junior employees have developed their abilities by undertaking routine tasks alongside more experienced colleagues. In offices, those tasks might include research, drafting, administration, reporting or preparing the first version of an analysis.
AI systems are increasingly capable of assisting with precisely this kind of work. That can be enormously valuable. Businesses can reduce unnecessary administration, improve services and give employees more time for work requiring judgement and human interaction.
We should welcome those opportunities.
But what happens to the training ground for experienced workers when the tasks traditionally performed by beginners can increasingly be automated?
It would be premature to claim that AI is already responsible for Britain's weak labour market. The evidence does not justify such a sweeping conclusion. Adoption varies considerably between employers and occupations, and the technology can create work as well as replace tasks.
The risk is more specific. Where a business is already uncertain about the cost of taking on an inexperienced employee, cheaper automation may tip the decision against recruitment. That would not necessarily produce dramatic announcements about job losses. It could appear gradually through fewer junior vacancies and narrower routes into established professions.
There is a further reason the Government cannot afford to treat AI and employment as separate policy questions: its own fiscal ambitions increasingly depend on productivity.
An October analysis by the Computer & Communications Industry Association, using OBR productivity assumptions and its own fiscal modelling, estimates that AI could account for approximately £4.8 billion of the Chancellor's £14.5 billion fiscal headroom in 2029–30.
The same analysis suggests that a more optimistic productivity path could improve fiscal headroom by around £35.5 billion compared with the central forecast.
Those figures are scenarios, not money already collected or guaranteed savings. The more ambitious estimates depend on assumptions about the speed and scale of productivity improvements, including how much of a broader productivity upside can reasonably be attributed to AI.
Nevertheless, they illustrate the potential prize. Even relatively modest improvements in productivity can have substantial consequences for tax revenues and the public finances.
That makes the Government's approach to AI more important, not less. Britain needs the technology to deliver real economic gains. But there is a danger in discussing those gains primarily as higher output and additional Treasury revenue without asking how businesses, workers and future entrants to the labour market will participate in them.
An economy can become more productive without automatically creating more opportunities for inexperienced workers. The distribution of the gains matters, as does the process through which people acquire the skills to contribute.
There is much in the Budget's technology programme that deserves support. Additional funding for sovereign AI capability, AI security and cyber resilience reflects legitimate national priorities. Britain should retain the ability to develop, deploy and scrutinise important technologies rather than become wholly dependent on decisions made elsewhere.
The Government has also recognised the challenge of youth employment. Its package includes a £3,000 Youth Jobs Grant for eligible hires, apprenticeship hiring payments for qualifying employers and £100 million over two years to expand local apprenticeship services.
These are sensible areas for intervention. It would be wrong to pretend ministers have ignored the problem.
The question is whether the measures are sufficiently joined up.
A grant may help an employer make a particular appointment, but it cannot by itself repair a business model in which the ongoing cost of employing and developing somebody is difficult to justify. An apprenticeship place is valuable only if an employer can provide meaningful work and supervision. An AI investment programme is successful for the wider economy only if its benefits reach beyond the companies receiving the funding.
We need a strategy that connects these decisions.
First, the Government should publish a cumulative assessment of the cost of employment, particularly for small businesses and entry-level roles. That assessment should examine taxation, statutory pay, regulation and administration together, rather than assuming separate policy impact assessments tell us everything we need to know.
Second, support for youth employment should reward sustained recruitment and genuine development. The objective must be productive careers, not temporary placements that disappear when a subsidy ends. Employers should have a reason to invest in beginners and a practical route to doing so.
Third, AI policy needs a much stronger labour-market component. We should measure changes in vacancies, tasks, occupational entry routes and training opportunities, distinguishing the effects of AI from wider economic conditions. Where technology changes how people acquire experience, employers, educators and government will need to develop alternative routes to competence.
Finally, productivity policy must be grounded in what businesses actually achieve. Helping a company adopt AI should not be judged solely by software purchased or employees trained. The relevant questions are whether the business becomes more productive, whether workers become more capable and whether that success supports sustainable growth.
None of this requires hostility to technology or indifference to workers' rights. A productive economy should be capable of paying better wages and maintaining fair employment standards.
But those ambitions depend on businesses being willing and able to employ people in the first place.
The Conservatives should make that connection central to their economic thinking. A credible alternative would combine proportionate regulation, competitive employment costs, incentives for workforce development and an ambitious approach to AI adoption.
Britain's challenge is not simply to produce more AI companies or train more people to use their products. It is to ensure that technological progress creates a labour market in which people can still enter, learn, advance and earn a good living.
The real test of this Budget will not be how much money ministers announce for artificial intelligence or how many employment initiatives they launch. It will be whether businesses create productive jobs and whether the next generation can build careers within them.
On that test, the Government has yet to offer a convincing answer.