Artificial intelligence is increasingly affecting economies and labour markets across the world. For the International Labour Organization, the central question is not whether AI will transform work, but how this transformation can be managed to advance decent work and social justice. As outlined in the G7 Action Plan for a human-centered development and use of safe, secure and trustworthy AI in the World of Work, adopted in 2024, G7 members need to address both the opportunities and challenges arising from adopting AI to develop future-proof and resilient labour markets that leave no one behind.
Beyond headlines purporting a ‘jobs apocalypse’, ILO evidence offers a more nuanced message: AI is more likely to transform jobs than to eliminate them. Although one in four workers globally is in an occupation with some degree of exposure to generative AI, only a smaller share of global jobs (3.3%) faces a high risk of automation. In most cases, AI is likely to affect specific tasks within jobs, reshaping how work is performed rather than replacing workers altogether. But, as AI capabilities grow rapidly, we need to continue to monitor the impacts of AI, particularly in its agentic form, not only on tasks but also more broadly on business processes and work organisation, which are likely to have more profound implications for jobs.
However as the technology evolves, we know this transformation will be uneven. Exposure to AI varies significantly across sectors, occupations and countries. Due to differences in the structure of the economy, high-income countries are more exposed than developing countries, but the differences are asymmetric. The latest research on these differences suggests that developing countries may experience the disruptive effects of genAI faster than its productivity benefits. Within countries, overrepresentation in more exposed jobs, such as clerical and administrative work, leaves women potentially more vulnerable to the automation effects of AI.
Uneven transformation: exposure, risk and labour market gaps
For enterprises, especially micro, small and medium-sized enterprises, the adoption of AI presents both opportunities and challenges. Larger firms are better positioned to leverage AI for productivity gains, while smaller and informal enterprises often face barriers related to finance, infrastructure and skills. Addressing these constraints is essential to ensure that AI contributes to broad-based economic growth and decent job creation. Without specific policy action, digital divides risk becoming structural fault lines.
Often missed in the current debates is the impact of AI on job quality in terms of its implications for wages, working conditions, labour protection and workers’ rights. On the positive side, AI can reduce exposure to hazardous tasks and improve occupational safety and health. At the same time, algorithmic management is increasingly used to recruit, monitor and evaluate workers. While these tools can improve efficiency and productivity, they can also intensify work, reduce autonomy and create new forms of surveillance. Safeguards are needed to eliminate biases and discrimination, particularly in hiring processes.
Digital and AI skills are increasingly important for jobs and workforce adaptability, although demand for these skills is materialising slowly in low- and middle-income countries. But employers do not seek these skills in isolation. ILO analysis of vacancy data shows that jobs demanding AI and digital skills, also require foundational digital and cognitive skills, as well as managerial and socio-emotional competencies. Developing broad bundles of skills is, therefore, essential for workers to benefit from this digital transformation.
Building a human-centred AI framework
Against this backdrop, the ILO’s vision is clear. The development and deployment of AI are shaped by a range of factors, including international labour standards, which play a critical role in ensuring that fundamental principles and rights of work are respected in the context of AI. Social dialogue is central to this approach. Engaging governments, employers and workers in shaping how AI is used in workplaces is essential for building trust, managing transitions and ensuring that technological change benefits all.
Policy responses must be comprehensive. They include strengthening social protection systems to support workers through transitions, investing in skilling and lifelong learning, and integrating AI into employment policies. Governance frameworks need to promote transparency, accountability and the protection of workers’ data. At the same time, efforts are needed to close the digital divide through investments in infrastructure, digital public goods and skilling.
The challenge of promoting a human-centred and inclusive adoption of AI is therefore not only technological, but also one of national policy and global cooperation. G7 members are urged to address both the opportunities and risks through a comprehensive policy approach spanning key areas, including employment, skills development, job quality and working conditions, in line with international labour standards. Effective responses require strong social dialogue and collective bargaining, underpinned by robust and up-to-date evidence on the implications of these technologies for decent work. These efforts depend on strengthened international cooperation to advance a human-centred approach to the adoption of AI, drawing on national insights and policy experience, with the G7 playing a critical role.
