Dealing with plagiarism & using AI in teaching

Dealing with plagiarism

The lecturers in the Department of Political Science make every effort to assess your individual performance in writing academic work fairly. Just that: Your individual achievements - not those of others. Our experience shows that it happens again and again that a few students submit written work that is not their own work, but plagiarised. This is not only unfair to fellow students who work hard for their seminar papers and theses, but it also violates the rules of academic decency.

We have therefore decided on the following procedure for dealing with seminar papers and theses and ask for your understanding:

In addition to the usual printed version, all seminar papers and final theses must also be uploaded as a digital PDF version in the respective GRIPS course. Please ensure that the wording of the digital version is identical to the printed version. Without exception, the same submission deadlines apply to the digital version of the papers as to the printed versions. For legal security, we also ask you to submit an affidavit (link to PDF) to the members of the teaching staff of our department for all seminar papers and theses in the sense of a declaration of credibility.

With this regulation, we want to prevent a few from gaining an advantage over the large number of honest students through dishonest behaviour. We hope this is in your interest.

Seminar papers and theses can be checked by the department using anti-plagiarism software on the basis of the digital version submitted. These programmes now have a very high hit rate, as they work with a combination of internet research and stylometric forensics (detection of style and wording breaks).

If it is established that a submitted work is demonstrably plagiarised in the sense defined below, this work is deemed to be insufficient. Section 22 Failure, Withdrawal, Deception, Violation of Regulations (valid since 11 March 2026) of the BA Examination Regulations for the Faculties of Philosophy at the University of Regensburg provides for the following regulation in paragraph 3:

The special provisions for the subject of political science (Section 53) have provided the following since 11 March 2026:

(4) Cheating

1. In deviation from Section 22 (3) of the BA Examination Regulations for the Faculties of Arts and Humanities at the University of Regensburg, the following regulation applies to examinations.

2. If, when preparing a written paper or bachelor's thesis, the candidate violates the obligation to write the paper independently and to identify all aids and sources or attempts to influence the result of an examination by other deception or use of unauthorised aids to their own or another's advantage, the paper will be graded as "insufficient" (5.0).

3. In serious cases, the examination board may decide that the grade awarded in accordance with sentence 2 shall be offset by 50% against the grade achieved in the resit attempt and thus be included in the overall grade of the Bachelor's examination in accordance with § 16 or that the candidate shall no longer be given the opportunity to resit in accordance with § 19 para. 3 sentence 1 and thus the Bachelor's examination shall be deemed to have been definitively failed.

Plagiarism (within the meaning of the resolution of the German Association of University Professors and Lecturers on Safeguarding Good Academic Practice (https://zenodo.org/records/14281892)) is defined as the fact that texts from third parties are copied in whole or in part, verbatim or almost verbatim, in seminar papers or theses and passed off as the student's own academic work. In this sense, plagiarism also exists if the text has been translated into a language other than that of the original. This definition does not, of course, apply to verbatim citations and literal citations in inverted commas which are labelled as such and the source is indicated.

Submitting plagiarism instead of an independently written seminar paper is not a "trivial offence", but constitutes a serious violation of basic scientific rules and the Copyright Act (external link, opens in a new window) (§ 23, 24 UrhG) and fulfils the criminal offence of deception (§ 263 para. 1 StGB (external link, opens in a new window)). It is also a criminal offence to submit a piece of work to obtain a certificate of achievement that has already been submitted in whole or in part in another course in political science or in another subject. Of course, we will exchange cases of plagiarism between the teaching staff of the Institute.

Dealing with AI in teaching

"Artificial intelligence (AI) is changing the way we learn, teach, research and work - including at universities. In a short space of time, AI-supported applications such as ChatGPT, DeepL Write and image generators have found their way into everyday study and work life. Many students, teaching staff and employees are faced with the challenge of dealing with these new technologies responsibly and competently" (Prof. Susanne Leist, vice-president for digitalization, networks and transfer).

The University of Regensburg has therefore drawn up a "Guide to the use of artificial intelligence for members of the UR"(https://www.uni-regensburg.de/universitaet/digitalisierung/leitfaden-zur-nutzung-von-ki) (external link, opens in a new window), taking into account the European Union's AI Regulation (KI-VO).

The European AI Regulation aims to ensure the safe, ethical and responsible use of AI in all areas(https://www.uni-regensburg.de/universitaet/digitalisierung/bedeutung-der-eu-ki-verordnung) (external link, opens in a new window) and to categorise its potential uses into different risk classes. The aim of the UR guidelines is to provide guidance on the following questions:

1. Where and how can Kl be used sensibly?
2. What is permitted, what should be critically considered - and where are the clear limits?

The guide aims to encourage "competent, creative and responsible use of Kl". At the same time, it aims to raise awareness of the legal, ethical and university-related framework conditions and sensitise all those involved to the fact that the use of AI brings both new opportunities and new challenges.

The Institute of Political Science follows the UR guidelines for dealing with AI in teaching. In the context of these guidelines, artificial intelligence (AI) isunderstood as "computer-aided systems that perform tasks that usually require human intelligence - such as understanding, processing and generating texts, recognising patterns or making (provisional) decisions." In particular, this involves the use of so-called generative AI, i.e. instruments or tools that can generate new content such as texts, images, code or language. To do this, it is necessary to train the AI systems by inputting a wide range of information, but this should relate to "content that is harmless in terms of data protection and copyright". This means that compliance with data protection and consideration of copyrights is essential. Examples of such tools are

• Voice-based AIs such as ChatGPT, Gemini or Claude
• Language and translation services such as DeepL
• Image and media generators such as DALL-E or Midjourney
• Kl-based assistance systems in Office applications such as Microsoft Copilot.

The use of these tools requires a high degree of responsibility, particularly in the academic sector, especially with regard to data protection, transparency and scientific integrity. The question therefore arises as to when the use of generative AI in university teaching is sensible and responsible Not only for the students, but also for the academic staff. According to the UNESCO International Institute for Higher Education's decision matrix, users should first answer the following questions:

1. Is it important that the content is correct?
2. Do I have the necessary technical expertise to assess the accuracy of the content of the Kl results?
3. Am I prepared to take responsibility for any errors?
4. Are there legal or ethical reasons (data protection, copyright) that speak against the use of Kl?

According to the guidelines, three different areas of use can be derived from this:

Area of use 1:

"Uncritically usable" is generative Kl e.g. to support creative work and to generate ideas. In other words, when initial drafts are created and then further developed manually, e.g:

• Brainstorming and idea development for projects or presentations
• Support in developing new concepts or strengthening existing ideas
• Creation of templates for quiz questions/teaching materials that are further processed Initial translations of teaching texts or communication materials into different languages; correction of spelling and grammar

Area of utilisation 2:

"Conditional use" is possible if the users are able to check the results and take responsibility for them based on their expertise. In order to be able to take responsibility for the results, it must be ruled out in particular that they violate personal rights, data privacy laws, licence and copyright laws or criminal laws. It should always be noted that even the provision of data (e.g. by uploading it to a Kl system) on servers that are not adequately protected by data protection or copyright laws may constitute a violation of applicable legal requirements (see compliance with data protection and copyright laws). Typical examples of use are

summarising articles for an initial overview of content or structure

• Analysing texts or data, e.g. to identify patterns or correlations Suggestions for formulations,
• Feedback on style and structure for better readability
• Individualisation of materials for specific learning needs
• Feedback on or simplification of programming and statistical evaluation code

Area of use 3:

"Not suitable" is the use of generative Kl in areas in which the results are not independently checked for accuracy and appropriateness or if the violation of (data protection or copyright) legal or ethical requirements cannot be ruled out. Since generative AI has no normative judgement, it cannot independently take moral, legal or institutional requirements into account. In addition, due to the black box nature of currently available AI solutions, there is a lack of transparency or traceability in the derivation of the results. This can lead to biased or discriminatory results that remain unrecognised and are incorporated into decisions, e.g. by reproducing existing prejudices in training data or disadvantaging certain groups. In this case, users can neither check the accuracy and appropriateness of the results nor take responsibility. Typical examples of applications in which the use of generative Kl must therefore be avoided are, for example

• Automated Kl-based assessments of examination results without human review (e.g. automatic grading of seminar papers or theses)
• Processing of personal, confidential or copyright-protected data (e.g. use of generative Kl to analyse non-released, copyright-protected teaching materials)
• Automated decisions in selection processes (e.g. applications, scholarships)

If the use of AI by a member of the University of Regensburg in the context of one of the two areas of use NBI or NB2 is assessed as uncritical and sensible, certain principles and guidelines have been formulated for the responsible use of AI: https: //www.uni-regensburg.de/universitaet/digitalisierung/leitfaden-zur-nutzung-von-ki/allgemeine-nutzungsrichtlinien-und-hinweise-fuer-den-einsatz-von-generativer-ki-an-der-universitaet-regensburg/verantwortungsvoller-einsatz-von-ki. (external link, opens in a new window) These must be observed at all times. The use of AI with regard to the 3rd area of use is excluded at the University of Regensburg.