Teaching & Learning
- Use of AI in coursework is generally not uniformly regulated across the university and depends on the specific module or instructor
- Some departments issue stricter local guidance requiring prior agreement or restricting AI-generated text in independently completed written work
- Students are told to check module descriptions and exam regulations, and if AI use is allowed they remain responsible for the content and must document that use according to academic conventions
There are no regulations under examination law regarding the use of AI in studies and teaching at TU Dresden. So far, there are only isolated regulations in module descriptions, but no general regulations in examination regulations.
Generally, AI can be used if use is permitted by the lecturer or examiner in a module. You can find out whether and how AI may be used in your case in the module description and in the corresponding examination regulations. If this is not regulated there, ask your lecturer or examiner.
If the use of AI is permitted, the independently completed performance and all its contents are still your own. You are responsible for the content and correctness of your work, which is why you should state if and how you used AI and make your use transparent using a suitable citation method according to the standards of academic work.
Die in Eigenständigkeit zu erbringende Prüfungsleistung muss in eigenen Worten formuliert werden. Ein nicht gekennzeichneter Einsatz von KI-generiertem Text im Rahmen einer Prüfungsleistung entspricht daher nicht den Anforderungen an eine Eigenständigkeitserklärung.
Der Einsatz von KI-Tools sollte allerdings immer im Vorhinein explizit mit der/dem Prüfenden besprochen und vereinbart werden.
- Faculty are advised to specify allowed aids clearly and may require declarations of AI use
- Some departmental guidance treats undisclosed AI-generated text in independently completed assessments as inconsistent with independence requirements
- For exams and assessments, TU Dresden states there is no university-wide examination-law rule yet; permitted use depends on module descriptions, exam regulations, or examiner decision
There are no regulations under examination law regarding the use of AI in studies and teaching at TU Dresden. So far, there are only isolated regulations in module descriptions, but no general regulations in examination regulations.
Generally, AI can be used if use is permitted by the lecturer or examiner in a module. You can find out whether and how AI may be used in your case in the module description and in the corresponding examination regulations. If this is not regulated there, ask your lecturer or examiner.
Lehrende sollten sich bei Prüfungen, die unter Zuhilfenahme von KI angefertigt werden können, Gedanken darüber machen, welche Art von Prüfungsleistung sie abprüfen wollen und wie sie dies am besten tun können. Dabei sollten sie prüfen, ob sie eigene Regeln zum Umgang mit KI festlegen und kommunizieren wollen.
Lehrende sollten außerdem die Anforderungen an die Eigenständigkeitserklärung den Umständen entsprechend anpassen. Es kann sinnvoll sein, Studierende explizit die Nutzung von KI offenlegen und die Eigenständigkeit mit Unterschrift versichern zu lassen.
Die in Eigenständigkeit zu erbringende Prüfungsleistung muss in eigenen Worten formuliert werden. Ein nicht gekennzeichneter Einsatz von KI-generiertem Text im Rahmen einer Prüfungsleistung entspricht daher nicht den Anforderungen an eine Eigenständigkeitserklärung.
- The university advises using AI as support rather than as a substitute for one’s own learning and writing process
- Students are warned that AI can produce incorrect or biased information and that they remain responsible for verifying results
- TU Dresden permits and recommends AI as a support tool for learning and study processes, such as generating ideas, structuring work, revision, and personalized feedback, while emphasizing critical checking of outputs
The use of AI in studies and teaching can be useful in many ways. AI can support personalized and competence-oriented learning, for example, through individually adapted learning paths and feedback. AI can also assist students and teachers in creating and revising texts.
By using AI assistants, students can, for example, receive support in brainstorming and formulating texts. But caution is also required here: the generated texts should always be checked critically and adapted to your own requirements.
As a learning aid, AI can be used to support self-directed learning by asking AI tools to explain concepts, suggest new perspectives, or formulate possible exam questions.
KI kann in verschiedenen Schritten des Schreibprozesses als assistierende Technik eingesetzt werden.
Wichtig ist dabei, sich genau zu überlegen, an welchen Stellen KI unterstützen kann, aber auch, inwiefern man sich durch ihren Einsatz gleichzeitig wichtige Lernschritte abnimmt.
Research
- TU Dresden allows AI as a support tool in scientific writing only under human responsibility and with transparent disclosure
- AI systems cannot be listed as authors, and researchers must critically review, verify, and take responsibility for all AI-assisted content
- Departmental and study guidance also states that AI use in academic writing should be documented and may be limited where independent formulation is required
Generative models can be used as tools in the process of scientific work and writing, but they can neither replace the creative process nor critical thinking, and they must not be listed as co-authors.
Authors are responsible for every aspect of their work and all final decisions. They are accountable for the correctness, integrity, and originality of their work. Any use of generative models and AI-assisted technologies in the writing process should be disclosed in a transparent way, following common standards and as required by stakeholders.
Wissenschaftlerinnen und Wissenschaftler müssen sicherstellen, dass der Einsatz von KI-Algorithmen und LLM nachvollziehbar und transparent gemacht und Angaben dazu, in welcher Form und in welchem Umfang KI genutzt wurde, offengelegt werden. Die Forschenden sind verantwortlich für die wissenschaftliche Integrität von Publikationen, insbesondere mit Bezug auf den Schutz des geistigen Eigentums und die Sicherung und Korrektheit der zitierten Quellen.
The use of generative AI in the writing process should be made transparent in an appropriate way, following the standards of academic writing.
If AI is allowed to be used in your case, there are some special features to consider when using AI in academic work and writing, and you should document your use of AI carefully.
- TU Dresden permits AI use in research data analysis only with careful scientific oversight
- The university also warns against uploading personal or confidential data to public AI systems
- Researchers must validate data quality, avoid introducing bias or distortion, and ensure that AI-generated or AI-processed outputs are reviewed against disciplinary standards
In many disciplines, AI algorithms are used to process and analyze data. As in all scientific methods, it is the responsibility of scientists to apply these methods according to state-of-the-art standards and to critically review and scrutinize the results.
This implies that scientists ensure the quality of data used for this purpose, and avoid introducing bias or distortions in AI-supported processing or analysis. Researchers are responsible for all outputs generated, such as texts, figures, images, and codes, and are liable for their use in publications.
Es ist zwingend notwendig, darauf zu achten, dass KI-generierte Daten bzw. die von KI unterstützte Datenauswertung nicht zu verzerrten wissenschaftlichen Ergebnissen führen.
Verwenden Sie grundsätzlich keine persönlichen oder vertraulichen Daten in öffentlichen KI-Tools.
- TU Dresden requires researchers to use AI in ways consistent with good scientific practice, transparency, and legal and ethical standards
- The ethics guidance also says risks must be weighed before using AI algorithms and legal requirements such as copyright and data protection must be observed
- Researchers remain fully responsible for scientific integrity, originality, correct citation, and protection of intellectual property, and they must disclose the form and extent of AI use
All researchers and scholars have to abide by legal and ethical standards and are accountable for any use of generative models or AI-assisted technologies in this context. Specific standards and regulations are to be observed if established by funders, publishers and scientific communities.
Researchers are fully responsible and liable for all outputs generated by AI-assisted technologies or generative models, and are accountable for every aspect of their work. This pertains also to contributions to applications for grants and proposals for meetings, conferences and events.
Wissenschaftlerinnen und Wissenschaftler müssen sicherstellen, dass der Einsatz von KI-Algorithmen und LLM nachvollziehbar und transparent gemacht und Angaben dazu, in welcher Form und in welchem Umfang KI genutzt wurde, offengelegt werden. Die Forschenden sind verantwortlich für die wissenschaftliche Integrität von Publikationen, insbesondere mit Bezug auf den Schutz des geistigen Eigentums und die Sicherung und Korrektheit der zitierten Quellen.
Vor Einsatz von KI-Algorithmen und Large Language Models in der Wissenschaft sollte sorgfältig geprüft werden, welche Risiken mit dem Einsatz potenziell verbunden sind.
Wissenschaftlerinnen und Wissenschaftler müssen bestehende rechtliche Standards beachten, wie den Datenschutz, die Wahrung von Urheberrechten oder den Schutz des geistigen Eigentums.
Academic Integrity
- The university provides examples of citation methods and notes that no single standard exists yet
- TU Dresden requires or strongly expects transparent disclosure of AI use wherever it is permitted, following the standards of academic work
- Students and researchers must indicate whether and how AI was used, and some faculties recommend or require explicit declarations in assessment contexts
If the use of AI is permitted, the independently completed performance and all its contents are still your own. You are responsible for the content and correctness of your work, which is why you should state if and how you used AI and make your use transparent using a suitable citation method according to the standards of academic work.
Any use of generative models and AI-assisted technologies in the writing process should be disclosed in a transparent way, following common standards and as required by stakeholders.
Wissenschaftlerinnen und Wissenschaftler müssen sicherstellen, dass der Einsatz von KI-Algorithmen und LLM nachvollziehbar und transparent gemacht und Angaben dazu, in welcher Form und in welchem Umfang KI genutzt wurde, offengelegt werden.
Lehrende sollten außerdem die Anforderungen an die Eigenständigkeitserklärung den Umständen entsprechend anpassen. Es kann sinnvoll sein, Studierende explizit die Nutzung von KI offenlegen und die Eigenständigkeit mit Unterschrift versichern zu lassen.
Although there is no generally accepted citation standard for AI in academic work yet, there are some recommendations and style guides on how to deal with AI-generated content and AI-assisted technologies in the writing process. The following publications can serve as a guide:
- Undisclosed AI-generated text may violate independence declarations where such declarations apply
- Suspected misconduct should instead be handled through existing examination-law procedures and review of the submitted work
- TU Dresden states that there is currently no reliable way to technically prove AI use and warns against relying on AI detectors as evidence
It cannot yet be proven beyond doubt whether a text has been generated with AI. AI detectors can provide indications here, but so far neither false positive nor false negative results can be ruled out.
Bei konkretem Verdacht auf einen Täuschungsversuch muss ein Prüfungsversuch auf Grundlage von §14 Abs. 1 AllgStPO geprüft werden. Bei Zulassung zur Prüfungsleistung ist zusätzlich §14 Abs. 4 AllgStPO zu beachten.
Eine automatisierte Einstufung einer Prüfungsleistung als Täuschungsversuch durch ein KI-Detektionsprogramm scheidet aus. Das Programm kann lediglich Verdachtsmomente liefern.
Die in Eigenständigkeit zu erbringende Prüfungsleistung muss in eigenen Worten formuliert werden. Ein nicht gekennzeichneter Einsatz von KI-generiertem Text im Rahmen einer Prüfungsleistung entspricht daher nicht den Anforderungen an eine Eigenständigkeitserklärung.
Institutional & Administrative
- The university also states that AI should not be used as a substitute for human pedagogical responsibility
- Guidance for instructors emphasizes clarifying permitted aids in assessments and adapting independence declarations where appropriate
- TU Dresden permits teachers to use AI for teaching support but expects them to reflect on risks, communicate rules for student use, and protect confidential information
AI can support teaching staff in the preparation and implementation of teaching. For example, AI can be used in the preparation of teaching materials or to create quizzes. AI can also support teachers in the supervision and feedback of students.
Lehrende sollten sich bei Prüfungen, die unter Zuhilfenahme von KI angefertigt werden können, Gedanken darüber machen, welche Art von Prüfungsleistung sie abprüfen wollen und wie sie dies am besten tun können. Dabei sollten sie prüfen, ob sie eigene Regeln zum Umgang mit KI festlegen und kommunizieren wollen.
Lehrende sollten außerdem die Anforderungen an die Eigenständigkeitserklärung den Umständen entsprechend anpassen. Es kann sinnvoll sein, Studierende explizit die Nutzung von KI offenlegen und die Eigenständigkeit mit Unterschrift versichern zu lassen.
Do not use any personal or confidential data in public AI tools.
The use of AI in studies and teaching should always be critically reflected upon and should not be used as a substitute for human interaction or pedagogical responsibility.
- This guidance applies across teaching, learning, and research contexts
- TU Dresden prohibits entering personal or confidential data into public AI tools and points users to a university-provided, data-protection-compliant service
- The university recommends using the browser-based TUD-Copilot and indicates that approved use depends on data protection and information security requirements
For TU Dresden members, TUD-Copilot offers a possibility to use AI in a data protection and information security-compliant way.
Do not use any personal or confidential data in public AI tools.
The web-based TUD-Copilot is recommended for use in studies and teaching.
Verwenden Sie grundsätzlich keine persönlichen oder vertraulichen Daten in öffentlichen KI-Tools.
- Ethics and good scientific practice bodies also issue university-level guidance on AI use in research
- Its governance approach emphasizes legally compliant, critically reflected, and beneficial use of AI, and it provides central contact and consultation points
- TU Dresden has an institutional AI initiative that frames AI use in teaching, research, transfer, and administration, supported by central information and advisory structures
Artificial intelligence (AI) is changing and influencing many areas of our lives. This also affects TU Dresden as an institution in research, teaching, transfer and administration. The use of AI opens up a wide range of opportunities, but also challenges us to use this technology responsibly and in a reflection-conscious manner.
TU Dresden aims to support and empower all members of the university in using AI profitably and responsibly.
This page provides information and support services on AI at TU Dresden. Here you will find information on using AI in studies and teaching, AI in administration and AI in research.
Sie haben Fragen oder Anliegen zum Thema künstliche Intelligenz an der TU Dresden? Hier finden Sie die passenden Kontaktpersonen und Beratungsangebote.
Vor Einsatz von KI-Algorithmen und Large Language Models in der Wissenschaft sollte sorgfältig geprüft werden, welche Risiken mit dem Einsatz potenziell verbunden sind.
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 TU Dresden's AI Policies
TU Dresden has defined AI policies in 11 of 12 categories, with an overall coverage score of 92%.
TU Dresden requires or strongly expects transparent disclosure of AI use wherever it is permitted, following the standards of academic work. Students and researchers must indicate whether and how AI was used, and some faculties recommend or require explicit declarations in assessment contexts. The university provides examples of citation methods and notes that no single standard exists yet.
TU Dresden states that there is currently no reliable way to technically prove AI use and warns against relying on AI detectors as evidence. Suspected misconduct should instead be handled through existing examination-law procedures and review of the submitted work. Undisclosed AI-generated text may violate independence declarations where such declarations apply.
TU Dresden prohibits entering personal or confidential data into public AI tools and points users to a university-provided, data-protection-compliant service. The university recommends using the browser-based TUD-Copilot and indicates that approved use depends on data protection and information security requirements. This guidance applies across teaching, learning, and research contexts.