Teaching & Learning
- The university does not set one uniform rule across all units
- Departmental policies vary: Philosophy forbids submitting papers wholly or partly written by generative AI, Study of Religion permits limited support uses but prohibits rewriting or uncritical AI-generated content, Economics permits generative AI in seminar and final papers with disclosure requirements, Development Studies requires signed declarations and allows examiners to restrict or ban AI, and German Linguistics requires listing any AI-based tools used
● Submitting a paper that is completely or in part written by GenAI is considered plagiarism and therefore explicitly forbidden.
The following guidelines apply to term papers, BA/MA theses, as well as smaller written assignments or presentations throughout your studies.
Examples for permissible and non-permissible uses of AI-based tools
Copyediting Using spellchecking, grammar tools and style recommendations critically in addition to reading and editing your text yourself multiple times and asking peers for feedback.
Having an AI-based tool rewrite all or parts of your thesis.
Grundsätzlich gilt, dass die Verwendung von Generative AI-Programmen in Seminar- und Abschlussarbeiten erlaubt ist.
Starting in the summer semester of 2026, all written work in this programme (term papers, essays, Master's theses, etc.) that is not produced under direct physical supervision must include the following declarations, which must be clearly attributable to the work and signed:
Lecturers may restrict the permitted use of AI, up to and including a complete ban on its use. This must be announced in the course materials.
Sofern ich weitere Hilfsmittel oder generative KI bzw. KI-basierte tools verwendet habe (z.B. ChatGPT, DeepL), habe ich nach dem Literaturverzeichnis ein entsprechendes Verzeichnis angefügt, in dem ich aufliste, für welche Zwecke diese eingesetzt wurden
- Assessment-related rules are defined in some programme-level documents rather than as a university-wide exam policy
- Development Studies ties false AI declarations to at least a failing grade and lets lecturers impose stricter AI limits through course materials; Economics frames AI use in seminar and final papers through examination practice and examiner-defined rules
Das wirft die Frage auf, wie seitens der Prüfungspraxis mit dieser Entwicklung umzugehen ist.
Grundsätzlich gilt, dass der Prüfer/die Prüferin einer Arbeit die Regeln festlegt, so dass es auch bzgl. der Verwendung von Generative AI-Programmen zu Abweichungen kommen kann, die auf der Website des entsprechenden Lehrstuhls zu finden sind.
I am aware that any violation of the contents of this declaration will result in at least a fail grade for the examination.
Lecturers may restrict the permitted use of AI, up to and including a complete ban on its use. This must be announced in the course materials.
- The university encourages students to build competence in using AI critically and responsibly, especially for learning and research-oriented support
- Institutional pages describe AI as helpful for research and for developing students' competencies, while departmental guidance allows limited uses such as checking understanding, background information, and grammar support, with critical review required
Wichtige Chancen liegen in der Kompetenzentwicklung der Studierenden. Sie können die Technologie für bessere Ergebnisse insbesondere in der Recherche nutzen. Diese Software kann ebenfalls dabei helfen, im Dialog mit der Software beispielsweise neue Forschungsfragen zu entwickeln und zukünftig mit vorhandenen Forschungsdatenbanken abzugleichen.
In addition, students are sensitized to working with artificial intelligence and practice the critical use of computer-generated texts. However, they must be guided in how to do this.
Dazu gehören auch Kompetenzen, um digitale Medien - insbesondere Künstliche Intelligenz - im jeweiligen Fachgebiet und in zugehörigen Berufsfeldern produktiv, reflektiert und verantwortungsvoll einsetzen zu können.
Lit review Using AI-tools to double check your understanding of certain aspects or background information. Critically reflect on the tool’s output and refer to suitable academic literature to further strengthen your interpretation.
● Using AI for grammatical check is permitted.
- Programming-related AI use is not defined in a university-wide teaching policy
- One programme-level document in Development Studies expressly includes programming as a use that must be declared when AI tools contributed to the work
Please state to which aspects of this assignment/ essay/ paper/ thesis AI-related tools and applications have contributed. Name the applications in your report. For example: “ChatGPT, DeepL and Claude.ai contributed to the following aspects of this thesis: literature research, programming, data analysis, review and editing, language improvement, visualization.”
Research
- Policies for academic writing and thesis/manuscript preparation vary by department
- Philosophy treats AI-written text in term papers as plagiarism and forbids it; Study of Religion allows limited support such as literature help and copyediting but prohibits having AI rewrite thesis text; Economics allows generative AI in seminar and final theses subject to disclosure; Development Studies requires reporting which parts of a paper or thesis used AI; German Linguistics requires a post-bibliography list of AI uses
● Submitting a paper that is completely or in part written by GenAI is considered plagiarism and therefore explicitly forbidden.
The following guidelines apply to term papers, BA/MA theses, as well as smaller written assignments or presentations throughout your studies.
Copyediting Using spellchecking, grammar tools and style recommendations critically in addition to reading and editing your text yourself multiple times and asking peers for feedback.
Having an AI-based tool rewrite all or parts of your thesis.
Grundsätzlich gilt, dass die Verwendung von Generative AI-Programmen in Seminar- und Abschlussarbeiten erlaubt ist.
Please state to which aspects of this assignment/ essay/ paper/ thesis AI-related tools and applications have contributed. Name the applications in your report.
Sofern ich weitere Hilfsmittel oder generative KI bzw. KI-basierte tools verwendet habe (z.B. ChatGPT, DeepL), habe ich nach dem Literaturverzeichnis ein entsprechendes Verzeichnis angefügt, in dem ich aufliste, für welche Zwecke diese eingesetzt wurden
- Research-data and analysis uses are addressed only in selected departmental documents
- Philosophy prohibits AI for empirical, computational, inferential, and statistical analyses in term papers, while Study of Religion strictly forbids uploading empirical research data to AI software, including for transcription or interview analysis, on confidentiality and ethics grounds; Development Studies requires disclosure if AI contributed to data analysis
Using AI for empirical work (e.g. econometric, computational, inferential, statistical, etc. analyses) is also prohibited.
It is always strictly forbidden to use any empirical data you collected during your research as input for AI-based software (e.g. uploading audio recordings for automatic transcription or interview transcripts for analysis). This would entail breaching confidentiality and the agreements you made with your interlocuters and thus violates research ethics standards.
Please state to which aspects of this assignment/ essay/ paper/ thesis AI-related tools and applications have contributed. Name the applications in your report. For example: “ChatGPT, DeepL and Claude.ai contributed to the following aspects of this thesis: literature research, programming, data analysis, review and editing, language improvement, visualization.”
- Research ethics and integrity expectations are explicitly stated in departmental guidance
- Study of Religion links AI use on collected empirical data to breaches of confidentiality and research ethics standards, Economics states that the rules are intended to counter scientific misconduct, and Development Studies treats undeclared or improper AI-assisted text generation as plagiarism and ties violations of the declaration to at least a fail
This would entail breaching confidentiality and the agreements you made with your interlocuters and thus violates research ethics standards.
Damit ergibt sich die Notwendigkeit Regeln zur Nutzung dieser Programme zu erstellen, insbesondere auch um der Gefahr des wissenschaftlichen Fehlverhaltens entgegenzuwirken.
Please note that copying AI-generated texts that are not based on your own (preparatory) work in their entirety will be considered plagiarism.
I am aware that any violation of the contents of this declaration will result in at least a fail grade for the examination.
Academic Integrity
- Disclosure requirements are clearly defined in several programme and department documents
- Economics requires AI-generated text passages to be identified and accompanied by a separate register of AI applications and prompts; German Linguistics requires a list after the bibliography showing what AI or other tools were used for; Development Studies mandates two signed declarations, including a statement of which parts of the work used AI and which applications contributed
Sofern Textteile von einer KI-Anwendung erzeugt worden sind, habe ich sie kenntlich gemacht und ein gesondertes „Verzeichnis über den Einsatz von KI-Anwendungen“ beigefügt, in dem ich im Hinblick auf jede wörtlich oder sinngemäß übernommene Textstelle darlege, welche KI-Anwendung zum Einsatz gekommen ist und wie ich sie genutzt habe. Zu diesem Zweck habe ich bei Nutzung einer Chatbot-Anwendung für jede betroffene Textstelle die Frage(n) dokumentiert, die ich an die Anwendung gerichtet habe.
Sofern ich weitere Hilfsmittel oder generative KI bzw. KI-basierte tools verwendet habe (z.B. ChatGPT, DeepL), habe ich nach dem Literaturverzeichnis ein entsprechendes Verzeichnis angefügt, in dem ich aufliste, für welche Zwecke diese eingesetzt wurden (bspw. zur Ideenfindung, Gliederungsentwurf, zur Übersetzung oder sprachlichen Optimierung).
Starting in the summer semester of 2026, all written work in this programme (term papers, essays, Master's theses, etc.) that is not produced under direct physical supervision must include the following declarations, which must be clearly attributable to the work and signed:
Please state to which aspects of this assignment/ essay/ paper/ thesis AI-related tools and applications have contributed. Name the applications in your report.
- Enforcement is addressed through plagiarism, misconduct, and declaration rules rather than a general AI-detection policy
- Philosophy states that AI-written papers are considered plagiarism, German Linguistics allows plagiarism checking with university-provided instruments, Economics treats breaches of the signed AI declaration as examination deception, and Development Studies states that violations lead to at least a fail grade
● Submitting a paper that is completely or in part written by GenAI is considered plagiarism and therefore explicitly forbidden.
Ich bin einverstanden, dass die Arbeit einer Plagiatsprüfung durch die von der UBT bereitgestellten Instrumente unterzogen werden kann.
Wird gegen den Inhalt dieser Erklärungen verstoßen, so liegt ein Täuschungsversuch vor.
I am aware that any violation of the contents of this declaration will result in at least a fail grade for the examination.
Institutional & Administrative
- Institutional pages state that lecturers need new expertise and that training formats are offered for all lecturers; the ITS page also references AI competence duties for personnel under the EU AI Act
- The university provides AI-related training and guidance for staff and lecturers, but the supplied sources do not set a detailed policy on faculty use for grading, feedback, recommendation letters, or administrative writing
In addition, students are sensitized to working with artificial intelligence and practice the critical use of computer-generated texts. However, they must be guided in how to do this. This requires a rethinking of teaching strategies and new expertise among lecturers.
Currently, the Centre for Teaching & Learning in Higher Education (ZHL) offers and continues to develop appropriate training formats for all lecturers.
Dort heißt es im Artikel 4: "Anbieter und Betreiber von KI-Systemen ergreifen Maßnahmen, um nach bestem Kräften sicherzustellen, dass ihr Personal und andere Personen, die in ihrem Auftrag mit dem Betrieb und der Nutzung von KI-Systemen befasst sind, über ausreichende KI-Kompetenz verfügen [...]"
Der Kurs richtet sich an Beschäftigte in der Hochschulverwaltung sowie an Wissenschaftler*innen und bietet konkrete Beispiele aus dem Verwaltungsbereich.
- The university identifies approved or institutionally supported AI access paths and sets data protection limits on use
- It provides access to Microsoft Copilot and Academic Cloud models through university logins, states that using Copilot with the university bt account brings specific data-protection assurances, warns against student use without that login, forbids staff's non-compliant service use, and says prompts should not include personal or confidential data
Die Universität Bayreuth ermöglicht es über bestehende Lizenzen sicher und für Sie kostenlos auf ChatGPT und DALL-E zuzugreifen.
Wenn Copilot so angemeldet nutzen, profitieren Sie von einer Einhaltung europäischer Datenschutzrichtlinien und des Urheberrechts.
Nur wenn Sie sich bei Copilot mit der dienstlichen Kennung (bt-Kennung) der Universität anmelden, greifen die besonderen Datenschutzvereinbarungen von Microsoft (gewerblicher/kommerziellen Datenschutz). Diese umfassen die Zusicherung, dass die von Ihnen eingegebenen Daten nicht zum weiteren Training des KI-Modells verwendet werden (Urheberrecht).
Wenn Sie ohne Kennung oder mit einer privaten Kennung auf Copilot (oder auch ChatGPT selbst) zugreifen, gilt diese Zusicherung nicht. Von einer solchen Nutzung muss daher für Studierende abgeraten werden, eine solche dienstliche Nutzung für Mitarbeitende ist nicht erlaubt.
The free Copilot product is not subject to the contractual and data protection conditions of the Microsoft 365 A3 campus license. “Copilot for Web/Edge” is subject to commercial data protection without a separate data processing agreement (DPA). Therefore, it is not recommended for use with personal and confidential data and is only suitable for limited use
Prompts should not contain any personal data.
Prompts must not disclose confidential content.
- The university has an institution-level strategic and governance stance focused on competence-building, responsible use, and early engagement with AI in teaching and administration
- Its teaching mission explicitly includes developing students' ability to use AI productively, reflectively, and responsibly, the university says it is addressing the technology early, and the ITS page frames implementation around AI competence obligations under the EU AI Act and university-provided training
Künstliche Intelligenz wie ChatGPT wird auch die Lehre an den Universitäten nachhaltig verändern. Auch die Universität Bayreuth ist sich dessen bewusst und setzt sich frühzeitig mit dem Thema auseinander.
Dazu gehören auch Kompetenzen, um digitale Medien - insbesondere Künstliche Intelligenz - im jeweiligen Fachgebiet und in zugehörigen Berufsfeldern produktiv, reflektiert und verantwortungsvoll einsetzen zu können.
Vor der Nutzung von KI/AI sollten Sie sich mit den Grundlagen und Rahmenbedingungen für den praktischen Einsatz vertraut machen. Seit dem 2. Februar 2025 ist der "EU AI Act" in Kraft.
Um dieser Schulungspflicht nachzukommen, bieten wir Ihnen entsprechende Online-Kurse an:
DocuMark: Responsible AI Use for Academic Integrity
Knowing your institution's AI policy is step one. DocuMark helps enforce it fairly by empowering universities to manage AI-generated content, prevent cheating, and support student writing through responsible AI use.
Common Questions About University of Bayreuth's AI Policies
University of Bayreuth has defined AI policies in 12 of 12 categories, with an overall coverage score of 100%.
Disclosure requirements are clearly defined in several programme and department documents. Economics requires AI-generated text passages to be identified and accompanied by a separate register of AI applications and prompts; German Linguistics requires a list after the bibliography showing what AI or other tools were used for; Development Studies mandates two signed declarations, including a statement of which parts of the work used AI and which applications contributed.
Enforcement is addressed through plagiarism, misconduct, and declaration rules rather than a general AI-detection policy. Philosophy states that AI-written papers are considered plagiarism, German Linguistics allows plagiarism checking with university-provided instruments, Economics treats breaches of the signed AI declaration as examination deception, and Development Studies states that violations lead to at least a fail grade.
The university identifies approved or institutionally supported AI access paths and sets data protection limits on use. It provides access to Microsoft Copilot and Academic Cloud models through university logins, states that using Copilot with the university bt account brings specific data-protection assurances, warns against student use without that login, forbids staff's non-compliant service use, and says prompts should not include personal or confidential data.