AI Policy: AI-Skill v1

The use of AI tools (such as ChatGPT, Copilot, Claude, Gemini, etc.) is not only permitted in this course but is an explicit learning objective. You should learn to use AI tools actively, purposefully, and reflectively. The specific regulations of this policy regarding documentation and academic integrity apply.

In addition to basic AI competencies, you should learn the following in this course:

• To assess the specific strengths and weaknesses of different AI models and operating modes (e.g., local models versus commercially offered models) for different tasks.
• To apply strategies for verifying and validating AI-generated information using scientific methods.
• To discuss the ethical implications of AI use in the field.
• To use AI tools to support scientific work processes (research, analysis, writing) in a targeted and reflective manner.

The following types of AI use are permitted as personal learning aids and do not need to be specifically labeled:

• Use for individual exam preparation (e.g., creating practice questions, summaries of your own notes, flashcards).
• Generating explanations of concepts or technical terms for your own understanding.
• Translating foreign language texts (e.g., academic articles) for your own understanding.
• Use in explicitly designated practice phases during the course (when guided by the instructor).
• Spell and grammar checking through standard software (e.g., in Word).
• Support with formatting and citation styles for written assignments.

The following types of AI use are permitted, provided they are transparently and traceably labeled according to the guidelines of this policy:

• Support with brainstorming, structuring, and outlining written work, homework, or presentations.
• Help with linguistic revision and improvement of your own texts (beyond mere spelling/grammar correction, e.g., style, phrasing suggestions).
• Generating examples, analogies, or alternative explanatory approaches to illustrate your own arguments.
• Brainstorming possible arguments, counterarguments, or perspectives for an analysis.
• Support with literature research (e.g., suggestions for search terms, summarizing abstracts, searching for sources with "Deep Research" functions of AI tools).
• Generation of code snippets or debugging assistance (if relevant to the subject).
• Creation of visualizations and graphics to support presentations or written work (e.g., through tools for automatic visualization of data or concepts).
• Transcription and summarization of audio/video sources.

The following types of AI use are expressly not permitted in this course:

• Submitting AI-generated content as your own work without adequate labeling – this may be considered an attempt to deceive.
• Use of AI during supervised examinations (exams, etc.), unless expressly permitted by the examination regulations or the instructor – this may be considered an attempt to deceive.
• Using AI to circumvent learning objectives that explicitly require independent critical thinking, creative problem-solving, analytical skills, or specific methodological competencies.
• Uploading copyrighted course materials or personal data to cloud-based AI systems without explicit permission (see Section 3).

When using AI services, observe data protection:

• Do not enter any personal data (e.g., names, student IDs, email addresses, discussion contributions) into external AI tools (neither your own nor those of other students or instructors).
• Be aware that many cloud-based AI services store your inputs and potentially use them to train their models.
• Where possible, use more privacy-friendly alternatives such as AI services provided by your university or AI models running locally on your computer, if possible.

Most course materials provided in this course are copyrighted. For the following, explicitly released materials, processing by AI services is permitted under conditions:

• Materials marked with the note "AI use permitted" in the learning management system.

All other materials may not be uploaded to AI services.

Label the use of AI systems in written submissions as follows:

1. "Documentation of AI Use" section in the appendix:

• Tools used: List of all AI tools used with name, provider, and version if applicable (e.g., "ChatGPT (OpenAI, model GPT-5)", "Claude (Anthropic)").
• Usage protocol (examples): Document exemplary requests (prompts) and the resulting AI responses that influenced your work. Briefly describe how you used or revised the response. (Scope by agreement with instructor, e.g., 2-3 examples).
• Reflection: Add a brief personal reflection (approx. 100-200 words): How did AI use influence your work process? What were the advantages, where were the challenges or limitations?

• For verbatim or slightly paraphrased adoption of AI-generated passages, label these directly in the text (e.g., with a footnote: "This paragraph is based on a suggestion from ChatGPT.").

• If you found literature sources with the help of AI tools (e.g., through "Deep Research" functions), mark this in the bibliography with an addition like "[via AI research]" or document it in the section on AI use.
• Carefully verify all AI-suggested sources for existence and correct bibliographic information before adopting them.

To maintain equal opportunity, this course does not assume that you have access to paid AI premium services. The learning objectives of the course can be achieved in the following ways:

• Using free basic versions of common AI providers (e.g., ChatGPT, Claude, Gemini).
• AI services provided by your university, if available.

AI competence is an explicit learning objective of this course. Both the process and the result of AI use are assessed:

• The prompt log with reflection is a mandatory component of every submission and is graded.
• The quality of AI use – prompt quality, critical review of results, independent further processing – is included in the assessment.
• A final portfolio documents AI use throughout the semester and is part of the examination.

AI systems make mistakes. Do not blindly rely on AI-generated information. Pay particular attention to:

• Factual errors ("hallucinations"): The AI invents facts, data, or events that sound plausible but are wrong.
• Invented sources: The AI cites sources (books, articles) that do not exist or whose content is incorrectly represented.
• Outdated knowledge: The AI's training data is not always current. Information may be outdated.
• Incomplete or one-sided presentation: Important aspects, perspectives, or counterarguments may be missing.
• Bias: The AI can reproduce unconscious biases from the training data.

• Verify central statements and facts using scientific sources (e.g., academic literature, databases).
• Verify citations and source references – does the source really exist? Is the statement accurate?
• Compare the answers of different AI tools or actively search for opposing positions.
• Be particularly skeptical of very specific or surprising statements.

For better results when using AI, the following strategies for formulating requests (prompts) can be helpful:

• Be specific: The more precise your question, the better the answer. Avoid vague requests.
• Provide context: Briefly explain the background or goal of your request (e.g., "I'm writing a term paper about X and need Y").
• Assign a role: Ask the AI to take on a specific role (e.g., "Explain it to me like I'm a first-year student", "Respond as a critical reviewer").
• Specify format using examples: Tell the AI how the answer should look (e.g., "List in bullet points", "Create a table", "Summarize in three sentences", "Answer like in this example").
• Work iteratively: Ask follow-up questions, request clarification, or ask for revision of the answer. Note that some AI tools tend toward so-called "sycophancy" – they agree with you even when you're wrong. Therefore, critically question when the AI unreservedly agrees with you, especially on complex or controversial topics.
• Formulate neutrally: Avoid suggestive questions that steer the answer in one direction (instead of "Is it true that X is good?", better "What are the pros and cons of X?").

If you have questions or uncertainties about the use of AI in this course or about the interpretation of this policy, please feel free to contact me:

• During my office hours
• By email or in the course forum
• During the course (we will schedule time for discussions).

• Your university's information resources on AI in teaching.
• Your university's didactic advisory services.

The instructors conducting this course also use AI tools, for example to create materials or support communication. In doing so, we always ensure compliance with data protection and copyright, as well as critical review of AI results.

We assure you that work you submit, personal data, or discussion contributions will not be uploaded to or entered into external, cloud-based AI systems by us for analysis.

When assessing examination performances (e.g., exams, term papers, presentations) that count toward your final grade, we do not use AI tools. Assessment is carried out exclusively by human examiners.