Artificial intelligence is no longer framed in Ireland as a distant productivity tool. A new report suggests it could reshape work much faster, with up to 7% of current jobs at risk in the short to medium term. The sharpest pressure may fall not on the lowest-paid workers, but on highly educated staff in better-paying roles. That reversal matters because it challenges a long-standing assumption about who technology hits first. The findings point to a labour market shift that could alter incomes, tax receipts and the balance of opportunity across the economy.
Why the AI risk now looks different
The research, from the Economic and Social Research Institute and the Department of Finance, presents a central scenario in which 7% of current jobs could be lost. That is roughly one in 14 jobs. The report says artificial intelligence adoption among Irish firms is likely to lead to job losses concentrated among highly educated workers, reflecting the exposure of high-skilled occupations to automation.
This is the key change in the latest assessment: the technology is not being treated as a threat only to routine manual or low-wage work. Instead, the study suggests that clerical, professional and technology roles may face some of the steepest disruption. In other words, the labour shock may land closest to the parts of the economy that have traditionally offered career progression to graduates.
Which jobs face the strongest exposure
The report breaks down the impact by occupation and shows wide variation. Clerical roles are among the most exposed, with projected losses of 18% for general and keyboard clerks, 15. 8% for numerical and material recording clerks, and 14. 6% for customer service clerks. Technology and professional roles are also heavily affected, with estimated losses of 13. 7% for ICT professionals, 11. 4% for business and administration professionals, and 10. 6% for ICT technicians.
By contrast, some sectors appear far less exposed. Health professionals show no projected job loss. Construction and manual trades are also among the least affected, including 0. 6% for building trades workers and 2% for electrical trades. Sales and service roles fall between those extremes, with estimated losses of 9% for sales workers and 6. 8% for personal services workers.
The pattern matters because it suggests artificial intelligence could widen labour-market disruption into occupations that have usually been considered resilient. Entry-level roles in financial services and law are singled out in the report’s wider framing as among the jobs most at risk, placing pressure on the traditional routes into work for many college graduates.
Income losses, household pressure and public finances
The report’s deeper warning is not just about employment numbers. It says job losses are likely to outweigh gains for many households, producing an average decline in household disposable income under widespread adoption. The largest average losses are expected in middle and higher income households. That is significant because these groups are also more likely to include workers in highly educated occupations exposed to AI technologies.
The study says wage gains for those who remain employed are expected to be modest but broadly shared, and not large enough to offset the average fall in income caused by job displacement. It also warns that if job losses are sizeable, falling tax receipts and higher welfare spending could create potentially large pressures on the public finances.
The report adds that the tax and welfare system absorbs most of the income loss for lower-income households, but it also notes that worsening inequality appears under every scenario it considers. If AI adoption produces a near-permanent reduction in income tax, the case for widening the tax base becomes stronger.
Expert warning on transition and inequality
Karina Doorley, one of the report’s authors, said the effect of artificial intelligence on the labour market and the distribution of income remains highly uncertain. She said a speedy digital transition would help minimise inequality effects. The report also points to the need for training and upskilling, arguing that Ireland’s strength in third-level education could help the country adapt if AI-complimentary skills are added to university and post-Leaving Cert courses.
The authors, Karina Doorley, Sorcha O’Connor, Richard O’Shea and Dora Tuda, present the study as a warning shot rather than a final verdict. Their central message is that the scale of disruption will depend on how quickly workers and firms adjust, and on whether policy can soften the distributional impact before it hardens into longer-term inequality.
What the findings could mean beyond Ireland
Although the report focuses on Ireland, its implications reach beyond one labour market. Economies that depend heavily on services, higher education and knowledge-based occupations may see similar tensions if firms adopt automation quickly. The Irish case is notable because the exposure appears to rise with education and earnings, rather than falling with them. That makes the report especially relevant for governments trying to plan for a transition that could lift productivity while still reducing total employment.
If the most exposed workers are also the most skilled, the policy challenge becomes more complex. Education alone may no longer guarantee insulation from disruption. The bigger question is whether workers, employers and the State can adapt fast enough to keep the gains from artificial intelligence from being concentrated in fewer hands. What happens if the future of work rewards productivity, but leaves the economy with fewer secure paths into the middle?
