AI in schools in Hesse

The state's AI tool

Since 1 October 2025, all schools in Hesse have had access to the AI chatbot AIS.chat in the Schulportal Hessen, the state's online school portal. The school activates it in the portal's administration area. Before teachers use it, they obtain a competence certificate, which is only available through courses run by the Hessian Teacher Academy.

Students use AIS.chat via links or QR codes that their teacher creates for lessons. Hesse's regional media centres also offer general AI training.

For your planning, this means that anyone who wants to use AIS.chat with a class first needs the school to activate it and their own competence certificate. Keep a record of which teachers already have the certificate, so that you know in which classes the tool can be used.

What the German states and the EU require

On 10 October 2024, the Bildungsministerkonferenz, the conference of education ministers of the German states, adopted a recommendation on dealing with AI in school education. In it, the states agree that a general ban on AI for work done at home is neither effective nor workable. This leads to agreements on how AI is used and how AI-generated content is labelled. According to the recommendation, assessing performance remains a task carried out solely by teachers.

In July 2023, the Hessian Ministry of Education published the guidance “Artificial intelligence (AI) in schools and teaching”. It describes possible uses, the legal framework and ethical questions, covers data protection and transparency and includes practical examples.

At EU level, Article 4 of the EU AI Act applies: anyone who uses AI professionally must take measures to support the development of AI literacy among their staff. This also applies to schools and education providers. No particular level is prescribed. You can read more in the article AI literacy under Article 4. AI systems intended to evaluate learning outcomes or to decide on admission to educational institutions are classed as high-risk AI under Annex III. Strict obligations apply to them from 2 December 2027.

Which data may go into an AI tool

Names, marks, photos and other personal data of students do not belong in any AI tool. This also applies to AIS.chat: entering personal data there is not permitted. The details are set out in the state's guidance and your school leadership's instructions.

A decision aid to pin up in the staffroom:

• If you are creating material that does not relate to individual learners, such as exercises, explanatory texts or board notes, then use a tool your school has approved.
• If a text contains names, marks, photos or other personal data of students, then it does not belong in any AI tool.
• If you are unsure whether a detail makes a person identifiable, then do not enter it and ask the school leadership.
• If learners are to work with AI themselves, then use AIS.chat via a link or QR code that you create once you have your competence certificate. Other tools are subject to the providers' age limits; read more in the article How parents can guide their children with AI.
• If a tool is meant to assess work or suggest marks, then the mark stays with the teacher. At most, AI can provide drafts of feedback.

A template for the school rule

A rule for the whole school answers four questions: When is AI allowed in lessons and when is it not? How do learners state that they have used AI? What counts as their own work? Who gives feedback and marks? The teacher always gives the mark.

The following draft is a starting point for reaching agreement at your school. Replace the details in square brackets and adapt it to your type of school.

• Tools: In lessons, we use AI tools that the school has approved. [Contact person] keeps the current list.
• Data: We do not enter names, marks, photos or other personal data of students into any AI tool.
• Permitted extent: Every task states how much AI is allowed, from “no AI” to “full AI use”.
• Disclosure: Anyone who has used AI states which tool it was, what it was used for (ideas, outline or correction) and what they did themselves.
• Own work: A learner's own work is what they have worked out themselves and can explain in conversation.
• Assessment: AI can prepare feedback. The teacher always gives the mark.
• Review: [Contact person] collects questions and experiences. We review these rules at the end of each school half-year.

The article Tasks that show students' own thinking describes how to set the permitted extent for each task and how to design tasks so that students' own thinking becomes visible.

Training teaching staff and recording what you do

Teachers in Hesse have several options. The Hessian Teacher Academy runs its own AI courses, including those leading to the AIS.chat competence certificate. The media centres offer general AI training. In addition, accredited courses from other providers are listed in the event catalogue at akkreditierung.hessen.de.

Which option fits depends on your goal:

• If teachers want to use AIS.chat with their classes, then they need the competence certificate from the Hessian Teacher Academy.
• If the teaching staff want to practise with their own teaching material and develop shared rules, then in-school training is a good fit, for example as an INSET day.
• If you only need a first overview for a staff conference, then a short input session is often enough, and training can follow later.

Keep a record of what your school does, so that you can show it if asked: who attended which training and when, for example with certificates of attendance, the current rule with its date and which AI tools are in use.

A roadmap for the school year

The steps work for a primary school as well as for a large comprehensive school; only the scale differs.

• Take stock: Run a short survey of staff on who uses which AI tools for what, and who already has the AIS.chat competence certificate.
• Assign responsibility: Name a contact person and decide whether AIS.chat will be activated at your school.
• Train: Plan training for the teaching staff, depending on your goal through the Teacher Academy, the media centres or accredited providers.
• Agree the rule: Have the draft rule discussed in the subject departments and agree it at your school, for example in the full staff conference.
• Involve learners and parents: Present the rule in the classes and explain it to parents, for example at a parents' evening.
• Review: After one half-year, gather feedback and adjust the rule and the list of tools.

An example: a comprehensive school with around 70 teachers starts after the summer holidays with a short staff survey. The steering group presents the results at a staff meeting, and an INSET day follows in the autumn. That is where the draft rule is written, which the subject departments add to by the end of the half-year. The rule is then agreed and presented in the classes and at the parents' evening.

Next steps

• Check whether AIS.chat is activated at your school and who has the competence certificate.
• Pin the data protection decision aid up in the staffroom.
• Copy the draft rule and pass it on to the steering group or the subject departments.
• Decide where training and rules will be documented.

If you want to involve teaching staff, learners and parents together, lindenbrook offers AI training for teachers accredited by the Hessian Teacher Academy, a project day from grade 8 upwards and a parents' evening, individually or as a programme for the whole school. The training ends with a draft of your school's AI rules. The page AI for education gives an overview.

Which of the three areas is still open at your school: the teaching staff, the rule or data protection?

Frequently asked questions

Sources

1. Schulportal Hessen: Training on AI and AIS.chat (in German)
2. AIS.chat: Questions and answers (in German)
3. Hessian Ministry of Education: Guidance on AI in schools and teaching, as of July 2023 (in German)
4. Hessian accreditation event catalogue: AI training from lindenbrook (in German)
5. EU AI Act, Regulation (EU) 2024/1689, consolidated version of 27 July 2026, EUR-Lex
6. mpfs, JIM study 2025 (in German)
7. KMK: Recommendation for education administrations on dealing with artificial intelligence in school education, resolution of 10 October 2024, PDF (in German)

Updated: 1 October 2026. This article is not legal advice.