One in four jobs globally could be transformed, not replaced, by generative artificial intelligence (GenAI), according to a new study by the International Labour Organisation (ILO) and Poland’s National Research Institute (NASK).
Released on May 20, the report, titled ‘Generative AI and Jobs: A Refined Global Index of Occupational Exposure’, offers an international assessment of GenAI’s impact on employment.
The study draws from nearly 30,000 occupational tasks, as the research combines AI-assisted analysis, expert evaluation, and harmonised ILO data to chart where and how GenAI might reshape the labour market.
However, the report stresses that transformation, not wholesale job loss, is the most probable outcome of this technological shift.
Read also: Women most at risk as Gen AI disrupts jobs — Report
What the report says
The report emphasises that the figures represent potential exposure, not guaranteed job displacement.
Practical barriers, such as infrastructure limitations and workforce skill gaps, mean that GenAI adoption will vary by region and industry.
Rather than replacing workers outright, the study suggests that GenAI will more often augment tasks requiring employees to adapt rather than exit their roles.
It calls for inclusive and proactive policymaking, with a strong emphasis on social dialogue among governments, employers, and trade unions to ensure fair outcomes during digital transitions.
“We moved beyond abstract modelling to construct a tool grounded in real-world occupations,” explained Pawel Gmyrek, ILO senior researcher and lead author. “By blending expert insight with GenAI scoring methods, we’ve delivered a replicable framework to help countries pinpoint risks and respond effectively.”
Key insights from the study
During the study, occupations were mapped by the degree to which they are affected by GenAI, enabling policymakers to distinguish between roles susceptible to full automation and those more likely to undergo task evolution.
Read also: Artificial Intelligence: Africa’s most powerful untapped tool for innovation and inclusion
Here are some of the key insights:
An estimated 25 per cent of global jobs fall within occupations potentially exposed to GenAI, with a higher share in high-income countries (34 per cent).
Women are more exposed: In high-income countries, roles at the greatest risk of automation account for 9.6 per cent of female employment, compared to 3.5 per cent for men.
Clerical jobs face the greatest exposure, given GenAI’s capabilities to automate routine office tasks. Meanwhile, exposure is also rising in highly digitised cognitive fields such as media, software development, and finance.
Despite GenAI’s growing potential, full automation remains unlikely as many roles will continue to demand human oversight or judgment, even when some tasks become more efficiently managed by AI.
The study underscores that policy decisions will shape the extent to which AI-driven job transformation maintains or enhances job quality.
“This index gives us a clear view of where GenAI might have the most significant impact, helping nations prepare and protect their workforces,” said Marek Troszyński, senior expert at NASK and co-author of the paper. “Our next phase involves applying this tool to Poland’s detailed labour force data.”
“We must cut through the noise surrounding AI,” said Janine Berg, senior economist at the ILO. “This tool provides both clarity and context, essential for preparing labour markets around the world for a just and inclusive digital transformation.”
The publication is the first in a planned series of ILO–NASK reports that will delve deeper into national labour dynamics and offer technical guidance for shaping effective GenAI strategies, particularly in developing economies.
