As a university, we are required to assess and evaluate the competences of all students. Therefore, we have to make sure that artefacts (text, images, code, etc.) that were submitted for evaluation in the context of various courses (lectures, exercises, labs) have actually been created by the respective student by using his or her own skills and competences. Given the rapid proliferation and availability of AI-powered tools, the Institute of Technical Informatics therefore applies the following general policy for its courses:
Students are generally allowed to use AI-powered tools and Large Language Models (LLMs) for the purpose of spell-checking, translations (e.g., DeepL), and stylistic/grammatical revision of their own, i.e., self-written text. However, students are responsible for the correctness and originality of the generated content, and must ensure compliance with the principles of good scientific practice as described in the following document: Safeguarding_Good_Scientific_Practice_Guidelines.pdf
Students are not allowed to use AI-powered tools and Large Language Models (LLMs) to generate code, to produce images, to generate text (sentences, paragraphs, sections), and to rephrase/paraphrase someone else's material. The generation of code/text/images by such tools/models is treated as plagiarism and leads to the same consequences. See the following document for a more detailed explanation.
The TU Graz Policy regarding plagiarism can be found in the following document: Statute_part_Plagiarism_-_Satzungsteil_Plagiat.pdf
Course instructors may decide to deviate from the above rules. If the use of AI-powered tools and Large Language Models (LLMs) is explicitly permitted by course instructors (e.g., as part of an assignment or task), the use of such tools must still be thoroughly documented in written form by the students by explicitly disclosing the utilized tools and how they were used to generate text/code/images/etc.