Use of AI in your study program

How can I use Artificial Intelligence (AI) responsibly? This page summarizes everything you need to know about using artificial intelligence in your studies: All binding rules at the University of Stuttgart, practical recommendations for tools, helpful learning resources, and answers to frequently asked questions.

The use of AI is permitted at the University of Stuttgart in accordance with defined guidelines

1. Use of AI is permitted

I am allowed to use AI systems at the University of Stuttgart within the framework of these principles.
My use of AI should support – not replace – my personal engagement with course content, my independent thinking, and the development of my academic skills.

2. Decision on the use of AI in examinations

The extent to which the use of AI systems is permitted or encouraged in examination performances is determined by the respective examiners. In case of doubt, I ask the examiners for clarification.

3. Responsibility for input

I am aware that I am responsible for all content I enter into an AI system. I do not enter any copyrighted material (text, images, etc.) into AI systems that store such content or use it for training purposes. I do not enter any confidential, internal, or personal data into AI systems unless explicitly authorized by a regulation of the University.

4. Responsibility for output

I am responsible for the use of any output generated by AI. I am aware that AI systems can produce incorrect or discriminatory content and may hallucinate information. Before using any AI-generated output, I carefully review it for accuracy, potential similarities with existing works (unintentional plagiarism), and compliance with legal and ethical standards.

5. Transparency and fairness in AI use

In consultation with the examiner, I clearly and transparently document and indicate where, how, and which AI systems I have used in order to ensure traceability. By adhering to these principles, I contribute to fairness in academic study and assessment.

6. University-operated AI systems and fair use

I preferentially use GDPR-compliant AI systems provided by the University. When using university-operated systems, I act fairly and in a resource-conscious manner to ensure equitable access for all users.

7. AI competencies

To strengthen my AI skills, I can take advantage of the University’s training and learning opportunities.

The university's data protection-compliant AI tool

Responsible AI (RAI) is the central, data protection-compliant AI tool at the University of Stuttgart. It offers functions such as text generation, rephrasing, idea generation, data analysis, and translation. The tool is accessible only in the university network or via VPN and offers two model options.

Access RAI Video tutorials RAI

Advantages

  • Your inputs will not be used for training future models.
  • Processing only on EU servers
  • Your inputs will not be linked to your personal identity.
  • Suitable for copyrighted content

Model options

Enhanced data protection, suitable for confidential content, with data stored only for short-term processing at the GWDG. These models are operated by the GWDG, a scientific institution in Germany, and funded by federal grants – available to you free of charge. Please note: Due to non-commercial operation, response times may occasionally be slightly longer.

High-performance, 30-day storage on Azure. These models run on Microsoft Azure in the EU. The university pays per usage unit (token) – usage is free for students.

Use the GWDG models if your inputs contain sensitive or personal information – for example:

  • Applications or letters of recommendation
  • Emails with personal data
  • Content from projects that include interviews, surveys, or case studies

Please keep the principle of fair use in mind: Use RAI responsibly – especially for tasks where the use of AI is genuinely meaningful. This helps to avoid unnecessary costs for the university and ensures the sustainable use of resources.

Using alternative AI tools

You may use AI tools other than RAI, provided that no sensitive, confidential, or personal data is entered. For your own protection, please disable the use of your inputs as training data for future models in these tools. You can learn how to set up a privacy-friendly configuration for, for example, ChatGPT, Perplexity, and Gemini in the following video tutorial. The use of external tools is at your own risk. 

Courses on offer

The University of Stuttgart offers voluntary courses to promote the responsible use of AI.

AI learning opportunities on offer at the University of Stuttgart
External learning opportunities

Frequently asked questions (FAQ)

That depends on the requirements of your course. If in doubt, ask the examiners.

The requirement and manner of disclosure vary depending on the field of study. Please check with your instructors/examiners regarding the preferred method of disclosure.

Whenever possible, use RAI and adhere to the principles for students. You may use other AI tools, provided that no sensitive, confidential, or personal data is entered. When using external tools, pay attention to the settings and, if possible, disable the use of your inputs as training data.

If you enter confidential, internal, or personal data into a freely accessible AI tool (e.g., ChatGPT Free, Gemini, Perplexity, Grammarly Free), you risk the following:

  • Storage of your input: Many providers store text inputs permanently and may use them internally.
  • Model training: Your input may be used to improve the AI model. This means that parts of your data could later be shown to other users in a similar form. Example: If you, for instance, use ChatGPT to rephrase a paragraph from your unfinished bachelor's thesis, it could later appear in a similar form in responses to other users – posing a risk to the originality of your academic work.
  • Linking to your account: If you are logged in, providers may associate your inputs with your email address, name, or other profile information.
  • No control over data sharing: With cloud-based tools, especially those outside the EU, you have no influence over who can view or process your data.

Advantages of the centrally provided AI tool RAI (Responsible AI) at the University of Stuttgart:

  • Hosting exclusively on EU servers – either at the GWDG (Germany) or via Microsoft Azure EU region.
  • You inputs are not used for training purposes – neither GWDG nor OpenAI via Azure uses your texts to train the models.
  • No personal attribution by external providers – your inputs are not linked to your identity.
  • Suitable for copyrighted content – you can also work with copyrighted materials.

RAI is designed with privacy in mind, allowing you to enter non-public or study-related content. This might include:

  • Outlines or draft texts for your term paper, project work, or your thesis.
  • Excerpts from unpublished teaching materials (e.g., slide decks, scripts).
  • Scientific questions or ideas that you do not yet want to share publicly.
  • Content containing copyrighted sections (e.g., text excerpts for analysis).
  • Personal reflections or case studies from courses.

These types of content should not be entered into AI tools when you are unsure whether the inputs will be used as training data for future models. This includes many freely available tools such as ChatGPT Free, Gemini, Perplexity, or similar services.

Tip: Use external tools like ChatGPT only for general, non-sensitive questions, for example:

  • “What is the difference between neural networks and decision trees?”
  • "Give me five arguments for AI in healthcare"
  • "Help me find a topic for a seminar presentation"

Contact

This image shows Lisa Schöllhammer

Lisa Schöllhammer

 

AI consultant

 

Technical Information and Communication Services (TIK)

To the top of the page