University of Hamburg - to research, to teach, to educate and form, to homepage
Professorship for Information Systems and Digital Innovation
Nucleus Professorship
for Information Systems
and Digital Innovation

Photo: Goettling

Generative AI tools (gAI) and their use for student assignments and theses

Generative AI (gAI) tools are powerful technologies that leverage advanced machine learning techniques and vast datasets to allow users to create or enhance content in various formats, such as text, images, audio, and video. Popular tools include ChatGPT for text, DALL-E for images, and DeepVoice for audio.

These tools can be used in the context of academic research and writing tasks, such as ideation, data analysis, literature review, thesis writing or editing. I is crucial to employ these tools judiciously, particularly in academic contexts, to uphold academic integrity and quality standards.

For orientation, the University of Hamburg provides a number of resources on the use of gAI in teaching and research:

In the context of final theses, dissertations, presentations, or essays, the chair for Information Systems and Digital Innovation provides eight concrete guidelines for the responsible use of gAI. These guidelines focus on three key principles: responsibility, agency, and transparency.

  1. Responsibility means that all dissertation and thesis work throughout all stages of the research and writing process as well as its outcomes remain the creative responsibility of the student. Accountability for all processes and outcomes, including ideas, arguments and text generation, rests fully with the student irrespective of any research or writing aids or implements.
  2. Agency means that all decisions to use or not to use gAI and any implications of such a choice rest fully with the student. gAI may be used as an enabling tool or implement to support robust research and writing practices but all choices, including the choice which gAI output to use and how, remain students’ decisions.
  3. Transparency means that the obligation is on students to transparently disclose their use of gAI tools for process and outcomes through appropriate citation practices.

Students of the University of Hamburg can now use UHH-GPT: https://uhhgpt.uni-hamburg.de/login.php 

8 Guidelines for the use of generative AI tools for theses and dissertations

  1. Consultation: All thesis projects are unique. gAI tools are evolving constantly and rapidly. Students should proactively discuss the use of gAI in the research and/or writing process with their supervisors in advance.
  2. Transparency: Students must include in their theses a clear specification regarding the scope and type of gAI use. It should be evident where ideas and arguments originate and how they were developed. Usage scenarios that need to be specified include... 

    I have used gAI tools…

    - Not at all
    - In the idea brainstorming process
    - When creating the thesis outline
    - To generate individual text passages
    - To generate a certain percentage of the entire text
    - Or developing or optimizing software source code
    - For proofreading, spellcheck or optimizing expressions
    - Other specific uses: ...

  3. Documentation: In addition, footnotes should be provided for sections of text that include GAI tool output. Corresponding prompts, the gAI tool used, and the date of use should be detailed in a separate appendix.
  4. Coherence: A thesis must be written coherently. AI use can sometimes generate shifts in writing style, overly specialized/superficial wording, or repetitive formulations. To maintain the individuality and coherence of your work, ensure the writing reflects your own style and you use your own formulations for the sections of your thesis.
  5. Liability for Errors: When using gAI tools, the sources used for generating responses or potential plagiarism cannot always be traced, and answers can be wrong or biased (AI hallucinations). Authors are fully responsible for the accuracy and verifiability of their arguments and sources, as with non-AI-generated academic works.
  6. Innovative Use: The innovative application of gAI systems may be a distinct contribution as part of a thesis project, provided it demonstrates significant skills in using gAI and constitutes a standalone research contribution (e.g., for complex coding tasks). Such applications may be evaluated as a distinct contribution of the thesis. In such a case, such contributions must match research questions or objectives of a thesis and should be described in terms of procedures, findings, implications, and limitations.
  7. Data Privacy: It is imperative to ensure that no sensitive, commercial, private or otherwise confidential data (e.g., regarding interviewed individuals or internal company information) is fed into gAI tools. gAI use must always adhere to relevant data protection and privacy guidelines.
  8. Grading: Adherence to the guidelines for the use of gAI tools for theses and dissertations will be reflected in the grading of the thesis, depending on the nature, scope, and complexity of a thesis project.

These guidelines underpin the grading of all theses. For further inquiries, please consult your thesis supervisor for explanations or clarifications.