TL;DR

Seventy people aged 16 to 21 have started a three-week AI course in north west England, with apprenticeships at local employers waiting for most of them at the end. Ministers intend to use the pilot to design a scheme covering all of England from next summer. The policy treats AI training as the remedy for a problem AI is widely suspected of causing.

What the three weeks contain

Recruits come from Wigan, Blackburn, Blackpool and neighbouring towns. The syllabus opens with constructing AI tools of their own, moves on to how employers put the technology to work, and closes on keeping a person in the loop to check whatever the software returns. Running alongside is a distinctly unglamorous curriculum: timekeeping, teamwork, problem-solving under pressure, professional communication, and the office systems every workplace assumes you already know. IN4 Group handles delivery on the ground. Heinz and JD Sports are among the employers holding apprenticeships open, together with smaller regional businesses.

The government has labelled it the TechFirst AI Apprenticeship Pathway, an extension of a £200m programme announced in summer 2025 that it says has reached more than 100,000 young people and financed 500 domestic PhD students. Participants will work with tools from Microsoft, OpenAI and Anthropic.

The number driving it

In figures published in May, the count of 16- to 24-year-olds neither working nor studying reached 1.01 million — the first time above a million since 2013. An interim report from Alan Milburn, released the same month, warned the share could reach one in six by 2031. Andy Burnham committed last month to bringing the total down without naming a date by which he expects to have done it.

Where the logic strains

Bouke Klein Teeselink of King’s College London called the approach sound in principle, on the reasoning that junior staff become more appealing to employers once they are more productive. He was less convinced three weeks achieves that, since staying current with these tools is continuous work rather than a course you finish.

That doubt lands on already thin ground. Resultsense reported yesterday that the evidence base for retraining schemes is weaker than the policies leaning on it assume, and on Tuesday the ILO put 6.1% of young workers in the roles most exposed to AI. Entry-level white-collar work — the drafting, checking and summarising handed to newcomers — is precisely what these tools absorb first.

Looking forward

The pilot’s real test is not the training but the destination: whether those apprenticeships still exist in three years, and whether employers who now automate junior tasks keep hiring juniors at all.