By 2030, at least 32% of higher education students should be enrolled in STEM fields.
That was one of the milestones set by the European Commission, early 2025, in its strategic plan dedicated to these disciplines. To entrench the targets at the EU level, it opened the path for dialogue with the 27 member states.
They validated most of the target indicators. They enrolled them in a European Council resolution. This launches the second cycle (2026-2030) of the European Cooperation Framework on Education and Training.
AI, on the periphery in 2021, at the center by 2026
The initial resolution – which launched the first cycle – dates back to February 2021.
From one to the other, the overarching aim remains a “European Education Area.” This should enable learners “to pursue studies at different stages of their lives and to seek employment across the [EU].”
However, the Covid crisis is no longer the focal point. The 2021 resolution, among other things, highlighted its revealing effects regarding the “digital divide and connectivity gaps within the member states.”
The AI dynamic runs inversely. Five years ago, the Council mentioned it only once, and within a political context (the Commission’s 2018 communication titled “A coordinated plan in the field of artificial intelligence”). This time, it underpins the roadmap, feeding most of the objectives that structure it.
A new objective dedicated to digital skills
Four of the five objectives from 2021 were kept in their broad outlines. They “absorbed” the remaining one, which lent itself to a cross-cutting approach (“support the ecological and digital transitions in education and training through them”).
The 27 added a sixth objective: “Develop digital skills and education for citizenship.” Their premise on digital skills is that they have become essential across almost all facets of society and the economy. They describe them as “keys” to exercising citizenship, employability, social inclusion, and well-being—while noting that they “greatly contribute to enhancing European competitiveness.”
The digital education should encompass AI and its use in an “ethical and responsible” manner, the Council specifies. It adds:
- Media literacy
- Digital data mastery and tools
- Critical thinking
- Digital well-being
- Combatting misinformation, disinformation, and cyberbullying
These elements are interdependent. Understanding AI can be a response to misinformation. Just as data and digital tool literacy can strengthen critical thinking.
The lever of micro-credentials
References to AI appear concerning the objective of excellence and attractiveness of VET (vocational education and training). In this area, the 27 note a particularly pressing need to strengthen STEM (science, technology, engineering and mathematics). Including upskilling of teachers and trainers.
Same observation for higher education. A lever mentioned repeatedly: micro-credentials, whose use could effectively translate upskilling and reskilling efforts.
Target indicators specific to STEM
Some target indicators defined in 2021 remain relevant for the second phase. For example, by 2030, lowering to below 15% the share of 15-year-olds with insufficient achievement in reading comprehension, mathematics, and science. More precisely, those who do not reach at least level 2 on the PISA scale.
Another maintained target: fewer than 15% of eighth-year students (age 13-14) with inadequate mastery of computer tools and information literacy. The calculation base: the IEA study (International Association for the Evaluation of Educational Achievement).
On this point, the 2026 resolution adds a target of at least 12% of students reaching PISA level 5 (out of 6 levels). It thus affirms, in parallel, those included in the European Commission’s STEM strategic plan. Specifically:
- In initial VET at the mid-level, at least 45% of students in STEM fields, including at least 20% women
- In higher education, at least 32% of students in STEM fields, including at least 40% women
- Within doctoral programs, at least 5% of students in the ICT domain, including at least one third women
AI-generated illustration
