Eindhoven University of Technology

1. The use of AI tools is allowed as an aid for general functionalities (teaching and study tool/assistant/input for own work) unless explicitly forbidden by the examiner (see point 3). General functionalities include brainstorming, gaining inspiration, summarizing general information, refining one's own work (e.g., language correction, language assistant), translation, and self-study/sparring partner (e.g., generating practice exam questions and answers). AI tools are not reliable scientific sources and the output must always be evaluated critically according to academic practices (e.g., as stated in the TU/e Code of Scientific Conduct; TU/e, 2019). Staff is always responsible for the educational activities they design and/or perform. Students are always responsible for the work they submit.

采集日期: 2026-10-07 · 行号: 11–11

存档哈希: 7c42eea90ece80b872e543e6a356792f71ea1228509d7eab48f1497b434e73be

2. When using GenAI functionalities (creation of new content, replacement of own work), complete statements about the use are required. The important distinction with the functionalities as mentioned under rule 1 is when GenAI partially replaces or outsources the student's own work and learning process. If a student uses GenAI in a manner other than meant under rule 1, it must be explicitly stated. This way, the teacher can provide more targeted feedback on the acquisition of academic practices and responsible use of tools. Standard scientific referencing methods (e.g., IEEE, APA, ACM) are taught and applied. Otherwise, a complete statement must at least include: 2.1. The name and version of the tool ## 2.2. Purpose and method of use The generation of quantitative and qualitative research data with GenAI is fundamentally prohibited, unless explicit consent is given by the examiner (see point 3). Note: People responsible for educational programs and courses may impose, additional requirements on the form and content of the statement, such as a more detailed statement of use, examples of entered prompts, reflection on reliability and bias, and verification of information.

采集日期: 2026-10-07 · 行号: 13–21

存档哈希: 7c42eea90ece80b872e543e6a356792f71ea1228509d7eab48f1497b434e73be

5. Take advantage of the positive functionalities of AI tools, but remain aware of risks and be a critical user. It is your own responsibility to use AI tools consciously, critically, and responsibly. AI tools offer many wonderful opportunities. However, the use of tools also comes with risks regarding the reliability of output (e.g., factual inaccuracies, biases, non-existent references), environment (e.g., energy consumption and cooling water usage), and data processing (e.g., violation of copyrights, NDA, intellectual property, and privacy, security, and storage of personal, corporate, and research data). Therefore, do not enter sensitive information or data. Follow the GDPR.

采集日期: 2026-10-07 · 行号: 37–37

存档哈希: 7c42eea90ece80b872e543e6a356792f71ea1228509d7eab48f1497b434e73be

7. Teachers remain responsible for the assessment of students and the content of the education. Teachers are encouraged to use tools in teaching and assessment, such as automated grading of multiple-choice exams based on pre-determined answers. However, automated decision-making/grading based on a GenAI model without human oversight over the assessment process is not permitted. The Examiner is legally responsible for conducting exams and determining the results.

采集日期: 2026-10-07 · 行号: 43–43

存档哈希: 7c42eea90ece80b872e543e6a356792f71ea1228509d7eab48f1497b434e73be

6. When the use of tools is mandatory in education, the processing and storage of personal data and information must be well regulated. When the TU/e tools do not suffice, a suitable alternative must be provided to students. For example, an open-source LLM that is hosted on a local server. Or a department may obtain a tool, and TU/e central negotiates a processing agreement with the supplier. Students may never be required to purchase a tool (or version with more functionalities) by their own means.

采集日期: 2026-10-07 · 行号: 39–39

存档哈希: 7c42eea90ece80b872e543e6a356792f71ea1228509d7eab48f1497b434e73be

## 4. Using AI tools counts as fraud if any of the following conditions are satisfied: 4.1. The submitted work is no longer sufficiently the student's own, in the sense that knowledge, insight, and skills as described in the learning outcomes cannot be assessed and tested. Delegating work to tools (or to someone else) to this degree is not allowed because it affects the core of academic practices (TU/e Regulations Examination Committee, 2024). The student must always take responsibility for verifying and analyzing information and for their own scientific substantiation. Teachers guide students in the education to understand this connection. 4.2. The student has not included a correct statement about the AI use. 4.3. The student used AI tools when the teacher had communicated that it was not allowed or only partially allowed. A correct statement of tool use does not change this. The definitions of fraud/plagiarism as described in the Regulations Examination Committee of the program apply here. In case of suspected fraud, the Examination Committee must always determine on a case-by-case basis whether fraud has actually occurred. Scores from AI detection tools do not count as sufficient evidence of fraud. In case of suspected fraud, an additional investigation may be conducted through an oral check. This is not an additional assessment moment. However, students must be informed of this possibility in advance. Check – if applicable - the department’s policy on AI and AI detection tool use.

采集日期: 2026-10-07 · 行号: 27–35

存档哈希: 7c42eea90ece80b872e543e6a356792f71ea1228509d7eab48f1497b434e73be

Faulty references can display the typical patterns of generative AI (e.g., missing or wrong publication data). It is considered the responsibility of the student (or group) to have checked their reference lists for accuracy, especially when the work has involved some form of generative AI. The presence of a reference list with the typical patterns of generative AI will be considered a Category 2 infringement. A repeat offence will be considered a Category 3 infringement. We consider this form of fraud as more than a Category 1 infringement (sloppiness) as there is a conscious decision made to use an unsuitable tool.

采集日期: 2026-10-07 · 行号: 93–93

存档哈希: 75c607b3802c04bab5612d32aaffc43158e08902174695bdc91be063894e862f

Fraud comprises in any case any intentional action by a student, or failure to act by a student, which makes it partially or completely impossible for the assessor to determine the individual student’s development of the overall competence of designing and vision on design. Fraud with respect to a curricular learning activity or assessment includes: Submitting work under one’s own name that has been done by others or copied from others wholly or partially: this includes copying word-for-word or paraphrasing the work of others without indicating that the words or underlying ideas belong to someone else (plagiarism); actively (assisting in) offering or assisting in offering one’s own work to others, who might then submit it as their own work; improper cooperation, that is two or more students jointly doing (part of) a curricular learning activity of which it is known in advance that individual work is required. If an assessor, lecturer or coach suspects fraud in relation to an assessment or a curricular learning activity, he or she will record this in writing as soon as possible and submit this to the Examination Committee. The student in question will be given the opportunity to add written comments to the report written by the assessor, lecturer or coach, respectively. It is then up to the Examination Committee to take any measures it considers appropriate in the case in question, which in the most extreme case could lead to a proposal by the Executive Board to terminate the student’s enrolment in the program indefinitely. See the Examination Regulations for more information about the Fraud and Plagiarism regulations and the sanction ladder including the different categories.

采集日期: 2026-10-07 · 行号: 86–88

存档哈希: 75c607b3802c04bab5612d32aaffc43158e08902174695bdc91be063894e862f

## 4. Using AI tools counts as fraud if any of the following conditions are satisfied: 4.1. The submitted work is no longer sufficiently the student's own, in the sense that knowledge, insight, and skills as described in the learning outcomes cannot be assessed and tested. Delegating work to tools (or to someone else) to this degree is not allowed because it affects the core of academic practices (TU/e Regulations Examination Committee, 2024). The student must always take responsibility for verifying and analyzing information and for their own scientific substantiation. Teachers guide students in the education to understand this connection. 4.2. The student has not included a correct statement about the AI use. 4.3. The student used AI tools when the teacher had communicated that it was not allowed or only partially allowed. A correct statement of tool use does not change this. The definitions of fraud/plagiarism as described in the Regulations Examination Committee of the program apply here. In case of suspected fraud, the Examination Committee must always determine on a case-by-case basis whether fraud has actually occurred. Scores from AI detection tools do not count as sufficient evidence of fraud. In case of suspected fraud, an additional investigation may be conducted through an oral check. This is not an additional assessment moment. However, students must be informed of this possibility in advance. Check – if applicable - the department’s policy on AI and AI detection tool use.

采集日期: 2026-10-07 · 行号: 27–35

存档哈希: 7c42eea90ece80b872e543e6a356792f71ea1228509d7eab48f1497b434e73be