AI skills have been packaged as the universal cure-all for the future, but the risk to human capital is great, argues Al Bird, who says that AI training divorced from the realities of work, and concentrated in the hands of the few, will never build true capability. Instead, the real task for L&D professionals is to design development that enables people to apply new skills in the context of their role
AI innovation has become the poster child of productivity. Over the next decade, it could add up to £47bn a year to the UK in productivity gains, says the government’s latest Industrial Strategy. The recent Growth & Skills report from Skills England echoes this, identifying AI and automation as “megatrends” shaping the future of every sector.
Beneath the excitement is an urgency. In the words of the Industrial Strategy, “the revolutionary potential of AI has barely begun to be realised”, and as everyone hurries to prepare for its full force, the folly is packaging “AI skills” as a universal cure-all for the country’s future.
Businesses will find they haven’t ‘done more with less’, they’ve done less with less – less imagination, less resilience, and less ability to navigate ambiguity”
Millions of pounds are already being poured into an AI-ready economy through government initiatives like the AI Opportunities Action Plan and TechFirst skills package, while the demand for “bolt-on” AI training and apprenticeships highlighted by Skills England captures the zeitgeist influencing today’s learning agendas.
While this is all positive, the near-exclusive spotlight on AI skills risks creating an illusion of progress. It channels training budgets and apprenticeship reforms into programmes that sound modern, but may in fact be hollowing out the very capabilities that the UK economy depends on. If we are to steam ahead with the AI skills revolution, it must be balanced with a few caveats.
Are we losing human capital in the race towards AI?
The unfortunate side-effect of all this trust in AI is that organisations using it will, by design, reduce headcounts. That is what efficiency looks like. As capability with these tools grows, tasks that once required teams of experienced professionals are streamlined into outputs that a smaller, tech-savvy workforce (supported by automation) can deliver.
AI and jobs
Salesforce cuts 4,000 customer service roles as AI takes over
TUC calls for ‘step change’ as half worry about AI impact on jobs
Businesses warned against reducing recruitment in favour of AI
While yes, AI is a tool for productivity, what gets skimmed over is that its impact, in practice, is cost reduction, usually through job cuts. That’s not necessarily cynical; it’s simply what happens when processes are automated. It does, however, raise an important question: if efficiency naturally “cuts the fat”, what happens to the human capital that carries judgment, creativity and depth of expertise?
This is where the L&D agenda collides with reality. AI pathways are being written into training programmes at pace, backed by the Growth and Skills Levy. But many younger recruits are already arriving in the workplace tech-savvy. The irony is that those who might benefit most from structured AI training (senior professionals who could pair decades of judgement with new digital fluency) are less likely to still be in the system.
The deeper danger is not that AI adoption eliminates jobs, but that in reshaping roles, it quietly hollows out human oversight. Experienced professionals are the ones who know when to question a model, when to ignore an algorithm, and how to sense the difference between derivative output and original thinking. Remove that layer, and businesses will find they haven’t “done more with less”, they’ve done less with less – less imagination, less resilience, and less ability to navigate ambiguity.
Towards smarter integration
None of this is an argument against AI. It must be part of every apprenticeship, every skills programme, every leadership curriculum. But it should be woven in, not “bolted on”, and there needs to be more intention behind what we’re building.
That you can run an apprenticeship to build AI capability is in itself a weird idea. What does an “AI apprenticeship” actually teach? If the model is simply to train people in prompt-writing or navigating a handful of popular tools, then we’re not developing knowledge, we’re running click-around training.
Skills England expresses what effective AI upskilling should actually look like, acknowledging that training opportunities will need to be “highly specialised and blended with more traditional sector knowledge.” That’s the direction we need to go in.
In finance, for example, employers need cloud engineers and machine-learning specialists who also understand regulation. In life sciences, employees must combine AI with scientific expertise in bioprocessing and personalised medicine. In other words, AI capability is valuable only when integrated with contextual knowledge.
The current momentum risks funnelling investment into basic AI literacy rather than building the hybrid expertise the economy actually needs. In treating AI as a standalone skillset, we forget that its transformative impact only emerges when it augments existing expertise. Without that, apprenticeships will produce technicians who can operate tools, but not professionals who can shape outcomes.
AI training divorced from the realities of work, and concentrated in the hands of the few, will never build true capability
The Growth and Skills Levy is a chance to design programmes that marry technical fluency with enduring capabilities: leadership, management, creative problem-solving, sector-specific judgment. That means resisting the temptation to chase quick wins – “AI apprenticeships” designed for visibility, not capability – and instead using the levy to build hybrid teams. Teams where seasoned professionals bring domain expertise, and newer recruits bring digital fluency, and both learn how to interrogate, contextualise and apply AI responsibly.
Building AI capacity without losing human depth
The Industrial Strategy’s ambition to make the UK an “AI maker rather than an AI taker” is the right one. But to achieve it, our skills policy cannot lean on blanket AI apprenticeships and short-term bandwagons. Instead, it should focus on hybrid development: building teams that blend digital tools with sectoral expertise, and weaving AI into broader programmes in leadership, management and professional development.
This is where the calibre of training and apprenticeship providers matters. AI training divorced from the realities of work, and concentrated in the hands of the few, will never build true capability. The real task for L&D professionals is to design development that enables people to apply new skills in the context of their role, because when experience stays in the room, capability strengthens rather than fragments.
Providers should be L&D’s most critical partners in striking this balance, helping translate policy into practice, and ensuring that AI becomes a force multiplier for human expertise rather than a substitute for it.
Sign up to our weekly round-up of HR news and guidance
Receive the Personnel Today Direct e-newsletter every Wednesday
