Vocational education and training (VET) plays a central role in equipping people with the skills needed for work, supporting productivity, inclusion and economic resilience. At the heart of VET systems lie curricula and qualifications. They translate labour market demand into structured learning outcomes, recognised credentials and training opportunities. Yet the processes through which these VET curricula and qualifications are developed remain largely invisible to learners and the public, despite being critical to the relevance, attractiveness and adaptability of VET systems.
These development processes are under growing pressure, with rapid labour market change, while the volume and complexity of curricula and qualifications continue to expand – partly reflecting trends towards modularisation and more tailored provision. Although VET development processes are typically rigorous, collaborative and grounded in stakeholder dialogue, they are also resource‑intensive and can struggle to keep pace with fast‑moving change. In this context, AI has emerged as a potential tool to support and improve these processes.
AI entered education and labour markets quickly, often before VET stakeholders had a shared understanding of what it can do, how it should be used, or how it should be governed. In VET development in particular, adoption remains cautious: stakeholders recognise AI’s potentials, but it also raises critical concerns. This OECD work, Developing VET with AI provides an evidence‑based and stakeholder-consulted snapshot in this area, drawing on OECD surveys of over 290 stakeholders across 25 countries, ten country case studies identifying thirty‑one concrete AI use cases (published in a separate report), and stakeholder dialogue across 80 stakeholders via interviews or workshops. The analysis shows that AI use in VET development is already being applied across key stages of VET development, from skills anticipation and consultation to drafting, validation and alignment. At the same time, the evidence confirms that AI is not a shortcut: the core foundations of VET development continue to underpin VET systems, and become even more critical when AI is used. The challenge is thus not whether to use AI, but how to integrate it transparently, securely and effectively within existing VET development processes.
Building on this evidence, this work distils five principles (whose details are published in a policy brief) for using AI in VET development, aligned with existing OECD, European Union and international AI frameworks but tailored to the VET context. It also sets out policy considerations to support countries move towards more effective, inclusive and secure AI use in VET development. In addition, it informs future national and international dialogue on the role of AI in VET and skills systems in a broader sense, including the 2026 European Union’s VET Strategy.
The OECD acknowledges the pioneering efforts of countries, institutions and stakeholders that are already testing AI in this domain, and will continue to build an evidence‑based understanding of how AI can support the full VET cycle – from design to delivery, assessment and outcomes.