Guidance for students on the use of generative AI in their studies
Policies on the use of generative AI vary across courses, programmes, departments and faculties. It is your responsibility to make sure you are familiar with the policy that applies to your studies.
About the guidance
This guidance outlines the principles that frame decisions on the use of generative AI in courses at the University of Gothenburg. It also clarifies students’ responsibility to comply with those decisions.
This guidance outlines information relevant to all students, such as:
- the principles underlying course policies on generative AI,
- the definition and limitations of generative AI, products licensed by the University, and how generative AI may affect learning,
- the rules and regulations governing the use of generative AI during examinations, and what is considered cheating.
Differences between courses
Decisions about the use of generative AI in university courses are based on three things:
- the course’s intended learning outcomes
- existing rules and regulations
- the University’s vision, values, and sustainability commitments.
Since learning outcomes vary across courses, programs, and departments, generative AI policies will vary too.
Course-specific decisions about generative AI
Decisions on the use of generative AI are made at the course level based on the course’s intended learning outcomes. The decisions must align with the University’s rules and regulations, mission and values, and its commitment to integrating sustainable development into all operations, including teaching.
Course-specific decisions about generative AI must be clearly communicated to students, together with an explanation of when the use of generative AI is permitted, restricted, or prohibited. Students are responsible for following these decisions and abiding by any course-specific limitations on the use of generative AI in coursework and examinations. If students are required to use generative AI in a course, all students must have access to the same licence.
Generative AI and the University’s vision and values
The University of Gothenburg is a government authority and operates within a framework of laws, regulations, and public responsibilities.
As a government authority, the University is required to work for sustainable development and provide evidence-based education that benefits society as a whole.
This means that the use of generative AI may be limited or prohibited in courses based on considerations related to learning, data protection and privacy, GDPR, ethics and sustainability. For more information, please see Vision and values on gu.se.
What is generative AI?
Generative AI refers to technologies that use probabilistic models to generate content by predicting likely outputs based on patterns in large datasets.
Generative AI technologies can generate different kinds of content, such as text, code, images, music, speech, and video. These technologies exist as both standalone applications, such as Copilot (Microsoft) and ChatGPT Edu (OpenAI), and integrated features in the software and digital environments used in university studies, such as Word, Excel and PowerPoint (Microsoft).
As AI becomes increasingly integrated into technologies used in higher education, students are likely to encounter generative AI in various contexts, such as when searching for information, processing data, reading and analysing text, or completing academic work.
Licensed products
If you, as a student, are permitted or required to use generative AI in a course, you should use licensed products procured by the University of Gothenburg, since these are provided under agreements governing how user data is processed and stored. At the time of writing, all students have access to a University-procured Copilot (Microsoft) licence. Some students may also receive access to a ChatGPT Edu (OpenAI) licence when enrolled in certain courses. As stated above, if students are required to use generative AI in a course, all students must have access to the same licence.
Copyright or sensitive information
Even when using a University-procured licence, students may not upload material that is protected by copyright or that contains personal data or other sensitive information.
Students should be aware that AI-generated content may closely resemble material created by humans. In such cases, AI-generated content may be subject to copyright, and students should not assume that they own or may freely use content generated by an AI product.
Furthermore, under current legislation, copyright protection only applies to works created by humans. This means that content generated by AI is generally not protected by copyright law.
Copyright may arise in works created with the assistance of AI when the human contribution is sufficiently original and substantial.
Limitations of generative AI technologies
If you, as a student, are permitted or required to use generative AI in a course, you should keep in mind that these technologies have limitations. For example, text-generating products predict statistically probable word sequences, which may contain factual inaccuracies, fabricated references, or misleading claims presented as facts.
Generative AI technologies are trained on large datasets that may include material collected from the internet, such as posts on social media, discussion forums, news sites, and other publicly available sources. Some perspectives are overrepresented in training data while others are underrepresented. English-language content and discourse originating from the Global North are particularly prominent. As a result, AI-generated content may reproduce or amplify the perspectives and values embedded in training data.
A key component of AI literacy is understanding the capabilities and limitations of generative AI technologies and being able to critically evaluate and manage risks such as bias, inaccuracies, and fabricated content. Given these limitations, generative AI should be understood as content-generating technologies rather than sources of information. When used for university studies, AI-generated content should be treated as material that may appear plausible but is not necessarily based on facts or reliable evidence.
Generative AI and learning
It is important to distinguish between learning how to use generative AI and using generative AI to achieve learning outcomes in university studies. Learning how to use generative AI involves understanding what generative AI is, how these technologies work, their limitations, ethical issues associated with their development and use, and strategies for critically evaluating AI-generated content. In contrast, using generative AI to achieve learning outcomes involves a different set of considerations and skills.
Learning cannot be delegated – it requires you, as a student, to process and understand the content you are studying yourself. If you consider using generative AI to support your studies, remember that developing knowledge and skills requires time and effort. Studying at university involves active engagement with course material, such as interpreting tasks, reading complex texts, developing and revising arguments, and solving problems. It is through this process of thinking, testing ideas, and revising one’s work that understanding and academic skills develop. If generative AI is used as a shortcut to complete coursework, important parts of the learning process can be lost. Students are responsible for reflecting on how the use of generative AI affects their own learning, understanding, and long-term development.
Generative AI and the rules for examinations
In the context of examinations, the University of Gothenburg considers generative AI a “tool” governed by the Rules for Studies at the First- and Second-Cycle Level. The rules especially relevant to the use of generative AI are listed below. Click on the text to read the explanation.
You can also read the entire document:
Generative AI detection products
The University considers AI-detection products unreliable and therefore does not accept AI-detection reports as proof that a student produced the work they submitted for examination.
Additional guidance
If you have questions about using generative AI for coursework and examinations, contact your course coordinator.