The sharp decline in entry-level hiring across advanced economies has widely been blamed on generative artificial intelligence quietly absorbing junior tasks. Research from LSE's Centre for Economic Performance by Peter Lambert and Yannick Schindler complicates that story. Their analysis of hiring data shows the fall in entry-level recruitment began after the Covid pandemic but before the widespread adoption of AI tools, which is awkward timing for the AI-did-it hypothesis.
The research attracted attention in the Financial Times, where the Data Points column explored an alternative culprit: remote working. If senior staff are no longer in the office, the informal apprenticeship through which juniors learn by observation and absorb tacit knowledge breaks down, and firms may simply find it less worthwhile to hire people who need that kind of development. The timing of hybrid work's rise fits the hiring data more closely than the AI adoption curve.
The distinction matters enormously for policy. If AI is destroying the bottom rung of the career ladder, the response involves regulation, retraining and perhaps slowing deployment. If remote working is the mechanism, the levers lie instead with firms' organisational choices about mentoring, office attendance and how early-career roles are structured. The researchers continued developing the work through 2026, with a fuller analysis of what broke the career ladder published later in the year.