As advances in artificial intelligence spread across OECD economies, new analysis warns that physical labour jobs face mounting disruption, with roughly one in four positions now classified at high risk of automation once AI is taken into account. Across OECD labour markets, occupations at high automation risk account for around 27 to 28 percent of employment, but the share differs sharply between countries. Nordic economies record less than one fifth of jobs in this category, while several Eastern European member states exceed 35 percent, revealing concentrated pockets of risk today.
OECD analysis indicates that rapid progress in AI is expanding automation potential beyond clerical and cognitive work into manual and physical tasks. Technologies combining robotics, machine vision and advanced control increasingly replicate manual and finger dexterity and can operate in cramped or obstructed spaces. This capability matters most for jobs organised around repetitive movements and standardised procedures, such as basic assembly, picking and packing, or materials handling.
As AI systems learn to coordinate both perception and movement, they can substitute for a larger share of physical task bundles that previously required human workers on site. Manufacturing provides early evidence of how these dynamics reshape workplaces. Case studies of AI adopting firms show that the skills made redundant are mainly manual, concentrated in production and processing tasks. Among manufacturing users who report automation, around half say some existing skills have become less valuable as AI systems take over routine operations.
At the same time, employers describe new demand for roles focused on developing, calibrating and maintaining AI models and robotic equipment. These positions require technical literacy, data skills and problem solving abilities rather than the manual competencies associated with factory work.
Workers in exposed sectors express concern about how AI adoption will affect security and pay. Survey evidence across OECD economies suggests that roughly three in five workers fear their jobs could be completely displaced by AI within the next ten years. In manufacturing, worries extend to earnings, with around two in five workers expecting wages to fall as automation spreads through production processes.
Yet current OECD labour market analysis finds that AI is mainly changing job content and the mix of required skills rather than eliminating numbers of positions outright. Occupational exposure patterns show that AI does not target only low paid physical labour. High skill white collar occupations, including business professionals, managers and science and engineering specialists, currently display the highest exposure to recent AI advances because their work relies on non routine cognitive tasks.
Moreover, managers play a crucial role in AI adoption leadership, helping to shape the narrative around AI’s value and ensuring alignment with business objectives.
