Clerical and high-tech jobs most exposed to AI, study finds

TL;DR:

  • An Engineers Ireland study estimates around 110,000 Irish jobs could be vulnerable to AI automation in the short to medium term.
  • Clerical workers and high-tech professionals are judged most at risk, drawing on ESRI research putting about 7% of workers in the danger zone.
  • Three-quarters of people surveyed fear AI will cause job losses, and 57% want tighter regulation.

A new study warns that the jobs most exposed to AI are not only routine clerical roles but also parts of the high-tech workforce. Engineers Ireland’s report, drawing on Economic and Social Research Institute figures, estimates roughly 110,000 jobs in the Republic could be at risk in the short to medium term — about 7% of workers, broadly in line with international estimates of 6–7%. For UK readers, the occupational pattern is more instructive than the Irish headcount: white-collar and technical work, not just manual labour, is in the frame.

Fear outpacing evidence of gains

The report captures a familiar gap between anxiety and measurable benefit. Three in four of the 1,000 people surveyed feared AI would lead to job losses, and 57% wanted governments to regulate it more tightly. Yet Engineers Ireland notes there is still little hard evidence of AI lifting productivity or the wider economy: it cites MIT research finding 95% of organisations reported “zero gains” from AI pilots, and US National Bureau of Economic Research data showing 69% of executives at adopting firms saw little impact, with usage averaging just 1.5 hours a week.

That mismatch — widespread fear, thin returns — echoes the UK picture, where workplace AI use has doubled but the gains cluster at the top and headline-grabbing cuts such as British American Tobacco’s 5,500 job losses sit alongside patchy evidence of transformation. The lesson for British policymakers and employers is not that mass displacement is imminent, but that the exposure is concentrated in exactly the office and technical roles many assumed were safe — and that public concern is running well ahead of demonstrated benefit.

Looking forward

Engineers Ireland’s director general framed the response in terms familiar to any UK skills debate: visible regulation, economic planning, and retraining support for those affected. The study is a reminder that adjacent labour markets are wrestling with the same question Britain is — how to prepare a workforce for AI whose disruptive potential is clearer than its productivity payoff. On current evidence, the safer bet is investing in skills and transition support now, rather than waiting for the gains to materialise.