Vocational Training

Trades that gain, trades that face pressure

The popular narrative of AI displacing all work is poorly calibrated for trades. An AI cannot wire a house, replace a pipe under a floor, or frame a wall — the physical judgment, dexterity, and real-world problem-solving required are not near-term AI capabilities. What AI can do is handle significant portions of the surrounding work.

For electricians: load calculations, compliance checking against the Building Code, job scheduling, and quote generation are increasingly AI-assisted or AI-automated. An electrician who can work with these tools is more productive per hour; one who cannot is slower and more expensive relative to peers. The net effect on the trade is probably an increase in the value of skilled practitioners and a reduction in administrative overhead — but it may also mean fewer jobs for less-skilled workers who did the supporting administrative work.

The picture is different for some white-collar-adjacent vocational roles. Architectural drafting, some inspection work, and routine compliance roles are under more direct pressure. These sit at the boundary of vocational and professional work, and the transition pathway is less obvious.

The key question for training providers is how quickly to embed AI-augmented workflow into trades qualifications, and whether to do so as an add-on or as a fundamental redesign of how the trade is taught.

Te Pūkenga and a system in transition

New Zealand's vocational education system has been through sustained disruption. The 2020 merger of 16 polytechnics and 9 industry training organisations into Te Pūkenga was the largest structural change in the sector in decades. The reform was explicitly aimed at addressing quality inconsistency, financial fragility, and poor outcomes for priority learners.

By 2024, aspects of that reform were being reconsidered — the scale of centralisation created its own problems, and there are live debates about the right level of regional and sectoral autonomy. This creates both an opportunity and a risk for AI readiness.

The opportunity: a system in active redesign is more able to embed AI literacy into qualifications and delivery models than a settled, resistant institutional structure. Design choices made now will shape the sector for years.

The risk: a sector managing structural upheaval may have limited bandwidth to simultaneously transform its educational content. Reform fatigue is real, and asking an organisation to reinvent itself while still consolidating a massive merger is asking a lot.

The micro-credential question

Micro-credentials — short, assessed, NZQA-recognised qualifications in specific competencies — are a potentially powerful instrument for rapidly upskilling workers in AI-adjacent skills. The framework exists; the question is how to use it.

Three broad models are being debated internationally:

Government-led and funded — the state defines the credentials that matter, funds their development, and subsidises learner access. Provides quality assurance and alignment with public priorities, but risks being slow to respond to employer needs and potentially funding credentials that the labour market doesn't value.

Employer-led — industry groups and employers define and co-fund credentials, with NZQA providing recognition. More responsive to actual labour market demand, but risks a patchwork of narrow credentials that don't transfer across employers, and may underinvest in skills with public goods value.

Learner-led with entitlement funding — workers receive a training credit they can spend on NZQA-listed credentials of their choice. Maximises individual agency, but evidence from Singapore's SkillsFuture (the most mature example) shows significant uptake of lifestyle and hobby credentials alongside genuinely job-relevant upskilling.

Australia's VET sector offers a cautionary tale: decades of mixed public-private delivery have produced significant quality variation and some well-documented scandals around low-quality providers accessing public funding. The design of quality assurance matters as much as the funding model.

Apprenticeship evolution

New Zealand's apprenticeship system pairs on-the-job training with off-job learning over multi-year programmes. AI changes both components.

On-job training now increasingly involves AI tools — whether the apprentice uses them, and whether the master tradesperson can teach their effective use, depends on the individual workplace. Some firms are ahead; many are not. There is no systematic mechanism to ensure AI tool competency is developed and assessed during an apprenticeship.

Off-job learning delivered by training providers has more scope for intentional design, but the same teacher-readiness bottleneck that applies in schools applies here: tutors who trained before these tools existed need development, and that development takes time and resources.

The question is whether AI readiness in apprenticeships should be mandated in qualification standards (which Industry Training Organisations influence), left to individual employer and provider discretion, or addressed through a targeted government programme. Each has precedent in existing vocational policy.