Published on

The EU Drew a Line on AI in Education: 'Technology Must Not Lead People'

On May 11, 2026, the European Union's Education and Youth Council made a quiet but consequential decision: it formally adopted conclusions on AI in education β€” marking the first time the EU has officially addressed the relationship between AI and teaching and learning as a matter of education policy. The message was not a technology endorsement. It was a correction. Education, the Council declared, must remain human at its centre.

Contents

  1. The EU Makes AI in Education an Official Policy Agenda
  2. Teachers Are Not "AI Users" β€” Redefining the Teaching Role
  3. Three Risks the EU Officially Named
  4. Inequality: When Technology Widens the Gap
  5. What the Council Asked Member States to Do
  6. August 2026: The EU AI Act's High-Risk Deadline Is Real

1. The EU Makes AI in Education an Official Policy Agenda

What makes these conclusions significant is that they are a first. This is the first occasion on which the relationship between AI and teaching and learning has been discussed as formal education policy at the EU level.

The Council's concern was clear: AI is moving rapidly into classrooms across the EU, but 27 member states are responding in 27 different ways. Finland has made coding and AI basics compulsory in primary school. Germany has banned certain AI tools over data protection concerns. France is piloting its own guidance framework. With no shared direction, the risk of fragmentation β€” and harm β€” was growing.

The central principle the Council embedded in its conclusions: human-centred. Technology should serve as a tool in education, not its purpose. And the person wielding that tool must be the teacher.

2. Teachers Are Not "AI Users" β€” Redefining the Teaching Role

The most emphatic element of the Council's conclusions is its redefinition of the teacher's role.

Much of the AI-in-education conversation has treated teachers primarily as consumers of technology β€” the question being how teachers can make use of AI tools. The EU pushed back squarely against this framing.

The Council defined teachers as guides, mentors, and critical thinkers who help students navigate an increasingly complex digital world. It went further, calling for teachers to have an opportunity to contribute to the design and evaluation of AI tools β€” not as afterthoughts or users in the feedback loop, but as genuine co-designers of the systems used in their classrooms.

This is not a symbolic statement. It is a direct challenge to EdTech companies: teachers should be treated as partners in development, not as end-users.

3. Three Risks the EU Officially Named

These conclusions stand out because they name AI's risks in education explicitly β€” a counterweight to the relentless optimism that has dominated much of the EdTech narrative.

β‘  Reduced autonomy and over-reliance on technology

When AI does the thinking, students may lose the capacity to think for themselves. European researchers have already begun observing a pattern: students who frequently use general-purpose AI tools complete tasks at higher rates, but the cognitive capacities those tasks were designed to build show measurable decline. Convenience carries a cost.

β‘‘ Bias, misinformation, and data protection risks

AI-generated content can embed hidden biases and present misinformation with plausible authority. If students absorb AI outputs without the critical skills to evaluate them, education can become a pipeline for reinforcing false beliefs rather than correcting them.

β‘’ Widening digital divides

The ability to access and meaningfully use AI tools is already distributed unequally across class, geography, and age. High-quality AI learning environments tend to arrive first at already well-resourced schools, while less-advantaged schools fall further behind.

4. Inequality: When Technology Widens the Gap

Digital inequality is a serious and persistent problem within the EU. Internet access, device availability, and teacher capacity to use AI tools vary significantly across and within countries.

The Council did not treat this as a simple hardware problem. The real divide, it recognised, is in the capacity to use AI meaningfully. Just as access to smartphones does not guarantee the ability to use them well, deploying AI tools in schools does not automatically narrow educational gaps β€” and may widen them if only better-resourced teachers and schools can leverage them effectively.

The conclusions explicitly called for equal access to digital resources for all learners, and named inclusion and fairness as non-negotiable core principles of AI in education policy.

5. What the Council Asked Member States to Do

The Council's practical requests to the 27 EU member states can be summarised as follows:

  • Strengthen teachers' AI and digital competences: Include AI literacy and critical usage in teacher training programmes
  • Promote education-specific AI tools: Develop and deploy tools designed for educational contexts, not repurposed from general-purpose AI
  • Address unequal access to digital resources: Implement policies to close the digital gap between schools, regions, and populations
  • Support the well-being of teachers and learners: Develop guidelines that account for the mental and physical impact of digital technology use on both teachers and students

6. August 2026: The EU AI Act's High-Risk Deadline Is Real

These conclusions are especially timely because they coincide with a significant regulatory milestone: from August 2026, the EU AI Act's requirements for high-risk AI systems take full effect.

Most consequential AI applications in education β€” systems used for student assessment, admissions decisions, and learning pathway recommendations β€” fall under the high-risk category. This means schools and EdTech vendors must meet stringent requirements for transparency, data protection, and human oversight.

Many institutions are not ready. The Council's conclusions are not just a principles document β€” they are a signal to member states and the industry that the window for preparation is closing, and the cost of non-compliance is real.

AI is entering classrooms faster than the policy framework around it can be built. The EU's message with these conclusions is unambiguous: the speed of technology is not the speed at which education must move. People lead, technology follows. And the people in education are teachers.

Sources

Discuss on Twitter β€’ View on GitHub
The EU Drew a Line on AI in Education: 'Technology Must Not Lead People' | MINSSAM.COM