Should schools ban AI?

No. Students use it anyway, so a ban is the worst answer. Better: clear rules per subject and level, teach students to use it critically, and train teachers first. The difference between an assignment in lower secondary and a final thesis at university level is too big for one blanket ban or one blanket permission.

How do you deal with AI and plagiarism?

Not with detection software: it lags behind and hits the best students with false positives. What does work is redesigning assignments: oral exams, presentations, in-class writing, process portfolios. Prove it not through 'this student wrote this essay', but through 'this student can think this through out loud in front of me'.

Where does AI save teachers the most time?

In the admin around teaching, not in an AI tutor. Preparing report card meetings (from an hour to ten minutes plus review), parent letters in the school's writing style (about 80 percent time saved), first-line feedback on draft essays and class-level test analysis. Start with what keeps teachers away from teaching.

What rules apply to AI and student data?

Education works with special GDPR categories: children's data, health data, social-emotional information. Cloud AI with student data is only allowed with a data processing agreement, EU-only storage and no training on student data. The AI Act classifies some educational AI (admissions, high-stakes assessment) as high-risk, with mandatory transparency and human supervision.

AI in education: the classroom of the future

Pretending pupils aren't using AI is the worst answer a school can give. They're using it. In the WhatsApp group, on the bike ride home, in the toilet during a test. So the question isn't whether AI is in your classroom, it's whether your teachers can handle it better than a fourteen-year-old who's been using ChatGPT as a homework buddy since primary school.

Meanwhile that same teacher is stuck on the real problem: not the tech, but the admin load. Lesson prep, differentiation, parent letters, report meetings, the student information system, standardised testing. That's where AI delivers results faster than in 'the classroom of the future'.

Opportunities for renewal

The most interesting opportunity isn't an AI tutor that replaces the teacher, it's everything that currently keeps the teacher away from teaching. A primary school teacher spends a few hours a week on admin and correspondence that a solid AI process handles in ten minutes. A secondary school mentor preparing report meetings for 28 pupils gains an entire afternoon.

Same story at pupil level. Personalised learning has sounded like a beckoning prospect since the nineties, but nobody had the time to build 30 practice routes for 30 pupils. With AI you build three variants in five minutes: a short one for those who get it, an extended one with more examples, a visual one with a diagram. Not future music; a year 7 teacher is doing this right now.

The human factor stays essential

The teacher isn't going away. Not in 2026, not in 2030. A class of 28 pupils needs someone who notices that Lieke is quieter today and that the group needs 20 minutes after PE to settle back down. No model does that.

What AI does do is give the teacher their attention back. Fewer lesson plans to type, more conversation at the desk. Not a romantic idea, a workload intervention.

AI in education: challenges and responsibilities

First challenge: digital literacy, and not just for pupils. A teacher who doesn't understand how a language model invents what it doesn't know can't explain to a pupil why that answer on the physics test was nonsense. Before you teach pupils anything about critical AI use, train your own team.

Second: plagiarism and assessment. You lose this fight if you bet on detection software, because it always lags and produces false positives that hit the strongest pupils hardest. What does work: redesigning assignments. Oral tests, presentations, in-class writing, process portfolios. Prove it not through 'this pupil wrote this essay', but through 'this pupil can walk me through this thinking out loud'.

Third: outdated knowledge. Models hallucinate with conviction. For exact sciences, don't put pupils on a chat model without a source check.

Fourth: privacy and data. Education works with sensitive pupil data, including social-emotional information. No consumer tool belongs anywhere near that.

The way forward

A school board that takes AI seriously does four things at once. One: train teachers, not with a one-off study day but with practical work sessions where they learn to speed up their own lesson prep. The moment a year 5 teacher experiences that their parent letters are ready in ten minutes is the moment AI is in.

Two: clear rules for pupil use, per subject and per level. No blanket ban and no blanket permission, because the gap between a project in year 8 and a final thesis in the fourth year of a bachelor's degree is huge.

Three: give critical thinking a more prominent place in the curriculum. Source evaluation, model knowledge, what is a hallucination, how do you check a claim. Not a standalone AI subject, woven through English, history, biology, everything.

Four: a curriculum that moves with the times. What's needed two years from now, you don't know yet. Build in that agility, otherwise you're back at the same conference table in 2028.

AI in education isn't about the classroom of the future, it's about today's workload.

Concrete AI applications for practice

Four things that already work at a school or higher-education institution today. Report meeting prep, where AI drafts a first summary of the pupil's progress based on test results and logbook notes, ready for review by the mentor. What used to take an hour now takes 10 minutes plus review time.

Parent communication, where generative AI writes standardised but personally-felt parent letters (about report grades or school trips) in the school's tone of voice. Teacher reviews and adjusts. 80 percent time saving.

First-line feedback on draft essays, where AI gives students immediate feedback on draft versions of their work so they can improve them before submitting the final version. Not as assessment (the teacher does that), but as a coach who helps with the writing process.

Test analysis at class and level, where AI dashboards surface patterns that teachers otherwise miss, such as which topics score poorly or which pupils get stuck on a sub-skill. And curriculum adaptation, where AI helps teaching teams design differentiating lesson assignments and alternative exercises for pupils with different prior knowledge.

Educational institutions work with special GDPR categories: children's data, health data (additional support needs), and in some cases sensitive social-emotional information. Cloud AI with pupil data is only allowed under strict conditions: a data processing agreement and data storage inside the EU. That's no longer optional in 2026, that's the floor.

The AI Act categorises some educational AI applications as "high-risk", such as systems that determine admission or automate important assessments. These applications require, among other things, mandatory transparency and human oversight. Build AI Act compliance in from day one, not as an afterthought. Retrofitting compliance always costs double and produces a tool nobody trusts anymore.

How a project runs in an educational context

A project runs in phases. First a short analysis to see which processes deliver the most time savings. Almost always that's parent communication or test analysis, not the shiny AI tutor the software vendor is pushing. Then a targeted pilot on one process to test whether it works for your school culture, because a secondary school team works differently from a higher-education programme committee, and that difference determines whether the tool sticks.

Only then do you scale what works and stop what doesn't fit. Where possible, plan key moments around the school holidays to minimise disruption. Honestly: if the pilot shows no time saving or quality improvement after three months, stop. Pushing through because there's already budget in it is exactly how schools end up with two lost AI years.

Conclusion

AI in education isn't about the classroom of the future, it's about today's workload. Start with what keeps your teachers from teaching. Build around that a tooling stack that's compliant and human-supervised. Assume pupils use AI and teach them to use it critically instead of banning it.

The schools leading the pack three years from now aren't the schools with the most expensive AI licence. They're the schools where the year 6 teacher is done with their parent letters by four o'clock on Friday afternoon, not half past six in the evening.

Curious what AI can do for your business?

Take the free AI Scan and find out in 1 minute.

Start the AI Scan →

Frequently asked questions

Should schools ban AI?: No. Students use it anyway, so a ban is the worst answer. Better: clear rules per subject and level, teach students to use it critically, and train teachers first. The difference between an assignment in lower secondary and a final thesis at university level is too big for one blanket ban or one blanket permission.
How do you deal with AI and plagiarism?: Not with detection software: it lags behind and hits the best students with false positives. What does work is redesigning assignments: oral exams, presentations, in-class writing, process portfolios. Prove it not through 'this student wrote this essay', but through 'this student can think this through out loud in front of me'.
Where does AI save teachers the most time?: In the admin around teaching, not in an AI tutor. Preparing report card meetings (from an hour to ten minutes plus review), parent letters in the school's writing style (about 80 percent time saved), first-line feedback on draft essays and class-level test analysis. Start with what keeps teachers away from teaching.
What rules apply to AI and student data?: Education works with special GDPR categories: children's data, health data, social-emotional information. Cloud AI with student data is only allowed with a data processing agreement, EU-only storage and no training on student data. The AI Act classifies some educational AI (admissions, high-stakes assessment) as high-risk, with mandatory transparency and human supervision.