1. Foundations7 lessons
  2. Inside the Model5 lessons
  3. How Learning Works5 lessons
  4. From Model to Assistant5 lessons
  5. AI literacy basicsLearning path · 6 lessons · with learning record

Teaching & training

For educators & trainers

Do you teach, train staff, or run a continuing-education course? You may use the lessons of KI einfach verstehen freely for that, paid courses included. This page shows how, what you need to credit, and where the limits of the content are.

What you may use

Texts and graphics are licensed under CC BY 4.0. You may copy, share, translate and adapt them for any purpose. The only condition: you credit the source, link the licence and indicate whether you made changes. The website's source code is under the MIT licence.

This is what the credit looks like:

KI einfach verstehen, “<Title>”, CC BY 4.0, <URL>

For example:

KI einfach verstehen, “Tokenizers: How Language Becomes Numbers”, CC BY 4.0, https://ki-einfach-verstehen.de/en/lessons/tokenizer-ids-vocabulary/

Use the title and address of the lesson the text or graphic comes from. If you changed something, add “adapted”. Not covered by the licence: the project's name, logo and wordmark, the fonts, quotations from other sources, and the merch designs. You may of course name the project, as long as it does not look like an endorsement or partnership.

How to use the content

  • Learning path for preparation or follow-up. The learning path “AI literacy basics” walks through the foundations in a fixed order. Your participants need no account; their progress stays in their own browser.
  • One lesson per session. Every lesson is a complete chapter with examples, graphics and often an interactive demo. It works without the others.
  • Recall moments as a warm-up. The questions at the end of each lesson are a good way to open the next session: answer from memory first, then discuss together. There is no grade.
  • Markdown export for your own material. With the lesson actions you copy or save every lesson as Markdown, for example for Moodle, ILIAS or your own handout.
  • Print for handouts. The print view leaves out navigation and controls; in the print dialog you can also save as a PDF.
  • Graphics on slides and worksheets. The credit belongs directly below the graphic. The menu on each graphic gives you its link and credit, ready to copy.
  • Glossary for reference. The glossary explains the technical terms briefly. It works well as accompanying material.

A learning record for your participants

Anyone who has read every lesson on the learning path and done its recall moments can create a learning record for themselves: with name, date, learning goals and scope (6 lessons, estimated reading time about 84 minutes). It is marked “Self-declaration — not verified” and carries no score. The name is only entered in the browser and stored nowhere.

To be honest: the learning path covers the technical foundations, nothing more. It does not cover legal questions such as obligations under the EU AI Act, data protection, or copyright; risks and limits of specific AI systems in your own work context; how to operate particular tools or write prompts.

Documenting AI literacy measures (Art. 4 EU AI Act)

Since 2 February 2025, Article 4 of the EU AI Act requires providers and deployers of AI systems to take measures so that their staff have a sufficient level of AI literacy. The regulation does not require a certificate. The European Commission notes that organisations can keep an internal record of trainings (AI literacy Q&A); the German Federal Network Agency recommends documenting the kind of measure, its content and time scope, and the participants.

The learning path can be one part of such measures. It covers basic terms, machine learning and language models from a technical point of view. Opportunities and risks, your organisation's role and the specific AI systems you use are not covered; you need additional measures for those.

A possible wording for your records, with placeholders in square brackets:

Measure to promote AI literacy (Art. 4 EU AI Act)

Kind of measure: self-study with the freely available learning path “AI literacy basics” by KI einfach verstehen (https://ki-einfach-verstehen.de/en/learning-path/ai-literacy-basics/), [with a follow-up discussion on … / without facilitation].

Content: technical foundations of AI models in 6 lessons: Program, Algorithm, Model Compared; Input and Output: What a Function Does; Tokenizers: How Language Becomes Numbers; Scalar, Vector, Matrix, Tensor: the Building Blocks of Numbers; Probability and Softmax: How a Model Decides; Parameters, Training vs. Inference, Hardware: How a Model Runs. Recall moments for self-assessment, not graded.

Time scope: estimated reading time about 84 minutes, plus recall moments; completed between [date] and [date].

Participants: [names or roles, or group and number].

Evidence: the participants' learning records (self-declaration, not verified), filed on [date] in [location].

Additional measures: [e.g. introduction to the AI systems in use, internal usage rules, risks in your own context].

Not legal advice. The wording is a suggestion. Whether your measures are sufficient overall depends on your AI systems, tasks and risks. If in doubt, have it checked legally.

Let the project know where you use the content

Do you use lessons, the learning path or graphics in teaching or training? A short message helps the project understand where the content is useful and what is still missing. The email is already prepared; just add what you want to share.

An institution is only named publicly if you explicitly allow it. Otherwise your details are only used to reply to you.