AI in your studies: What is permitted?
The University of Vienna’s position
The University of Vienna supports the responsible use of artificial intelligence in your studies, provided that it is used transparently, in accordance with the rules and with scientific integrity, and that it helps you to achieve your academic goals.
General framework for the use of AI
The use of AI during your studies is not strictly prohibited, but nor is it permitted without restriction. Whether and how you are permitted to use AI always depends on the specific context, in particular on:
- the type of course
- the format of the examination or assignment
- the lecturer’s specific instructions (please do ask!)
- the relevant learning objectives
- the rules of good academic practice
Learning objectives and independence
Courses and examinations at the University of Vienna are designed to help you achieve clearly defined learning objectives. These aim at building subject-specific skills and teaching students how to conduct academic work. This includes understanding and presenting content accurately, applying knowledge to your own research questions, formulating arguments independently, structuring texts effectively, and using sources appropriately.
You need to acquire these knowledge and skills independently to achieve the learning goals.
If an AI takes on these tasks, the learning objectives cannot be achieved. The skills to be acquired through the learning objectives are demonstrated in various course and examination formats, such as examinations, presentations or written assignments. These skills must be demonstrated by yourself and not by an AI tools. Your work must reflect your own thoughts and learning processes.
This, then, also determines how AI may generally be used during a degree course:
-
AI can help with learning by, for example, helping you organise your thoughts or by improving your language skills.
-
However, it must not cover those areas of competence that are to be assessed as part of a course or examination.
How AI tools can be used therefore depends largely on the specific learning objective, and these can vary from one course to another.
Example:
- In an academic paper on a specific topic, you are translating a passage from French in order to analyse its specific content. In this case, the use of an AI translation tool may be acceptable, provided that the translation is checked for accuracy and the use of AI is disclosed.
The situation is different in a language degree, where translation itself is a key learning objective. If you were to hand over a translation task to an AI tool in this context, it would undermine the independence of the work that is expected of you.
If AI carries out your work for you, this contravenes academic integrity and the principle of independent work. Whether you are permitted to use artificial intelligence during your studies therefore depends primarily on whether this ensures that the required learning objectives are met and that your work remains independent.
To sum up:
AI may provide support, but it must not carry out the work that is expected of you.
This recommendation can be put into practice by specifically designing an AI tool to tailor the support provided to the relevant learning objectives. To this end, a kind of AI tutor can be developed which provides support whilst at the same time helping learners to maintain their independence.
What is permitted? And what isn’t?
Whether you can use AI in a module or during an exam depends on the learning objectives, the basic principles of good academic practice and the guidelines set by the respective lecturer. Therefore, you should always check the following:
- the details on u:find
- the information on Moodle, and
- details regarding specific assignments and exams.
If the use of AI is not clearly regulated there, clarify this with the lecturer in advance. If you are unsure, assume that AI is not permitted unless you have been given explicit permission.
-
This is usually permitted:
You may use AI, provided that you are responsible for the content and the originality of your work. You may use AI:
- to generate ideas or structure topics
- to gain an overview of a new topic or to prepare for reading
- to revise the language of your own texts
- to assist with reflection or to check for clarity
- to find examples or alternatives which you then review and evaluate yourself
It is always a prerequisite that the use of AI complies with the lecturer's instructions and does not replace your independent work.
-
This is not permitted:
The use of AI is not permitted if it undermines the required skills or if its use is not disclosed. It is therefore not permitted to
- submit AI-generated texts or solutions as your own work
- allow AI to carry out tasks, analyses or arguments entirely or for the most part
- use AI in exams or assessments where it has not been clearly permitted
- use AI-generated content without acknowledging it
- use Ai generated sources, quotations or bibliographies without checking them
Such a practice constitutes a breach of the rules of Good scientific practice and may be regarded as cheating.
Data Protection and Copyright
In addition to the guidelines derived from the learning objectives, the basic principles of good academic practice and the instructions provided by the lecturer, regulations on data protection and copyright are also relevant when working with AI.
-
Copyright
AI-generated content is not protected by copyright, but you must not present it as if it's your own academic work. There are also a number of things to bear in mind when using AI tools to create texts.
-
Data Protection
When you use AI tools in your studies, you need to be careful with personal or confidential data. It is not always clear how the content you enter is stored or processed.
-
Learning new content
AI summaries can provide an overview, but they are no substitute for engaging with texts independently.
-
Tasks that require absolute precision
Current AI models operate on the basis of probabilities and are therefore not suitable for tasks that require precision, such as exact calculations.
-
Tasks for which AI is unsuitable
AI performs surprisingly poorly on certain tasks and can even produce errors that are not apparent at first glance.
-
When the process is important
The only way to learn to write is by writing. This section explains why this process cannot be outsourced to AI, and how AI can nevertheless provide support.