When using AI in research, we critically examine the algorithmic fairness of methods, data, and results. 

We are aware of potential biases—for example, in the training data and, consequently, in the results—and understand that this awareness is a fundamental prerequisite for the responsible use of AI. We handle data with care and keep in mind that the data entered is usually reused by AI tools for training and personalization. This means, for example, that we do not upload any personal or confidential data to AI systems unless it is guaranteed that the data will not be stored there, reused, or processed for other purposes. When in doubt, we use local models that comply with data protection regulations or those with zero data retention. We adhere to the terms and conditions of the software providers.

We are aware that AI can infringe on copyrights during data mining and content creation. If AI-generated results are adopted without critical review, this can result in copyright infringement and, in some cases, plagiarism. Therefore, we take our responsibility as authors for the content of the publication seriously. For questions regarding the use of AI in research, we are supported by the Data Protection Officer [DE] and the Ethics Committee.

We understand that AI technologies operate on a probability-based basis. The results of AI-generated responses are not reproducible.

We take responsibility for our research data and observe ethical, data protection, and confidentiality considerations [DE] in our research data management practices.

In academic writing, we as authors take responsibility for the text. We disclose the use of AI in the creation of text, figures, and tables, as well as in the editing of texts or similar activities. Transparency is important to us. We cite sources in accordance with discipline-specific guidelines; when using AI, we cite it as a source with the standard citation information. We prefer to cite original sources rather than the AI itself. When using AI as a method, we describe this in the relevant section of the text and, if applicable, also specify the prompt used.

Many publishers already have guidelines regarding the use of AI and the citation of AI usage. We adhere to these guidelines.

Doctoral dissertations and habilitation theses represent a unique aspect of scholarly work: the goal here is to demonstrate independent scholarly work. AI-generated texts do not constitute independent work, and we do not use them without careful consideration.

Writing expert reports is a core task for us as scientists, one that requires our professional expertise. We can use AI as a supportive tool, for example, to convert keywords we’ve written ourselves into continuous text or to search for additional information. We do not delegate the evaluation itself to AI. Furthermore, we maintain confidentiality. We do not upload grant applications or manuscripts to AI tools unless it is guaranteed that the information will be treated confidentially and that the data will not be stored and/or used for other purposes. As reviewers, we take full responsibility for the review.

External funding agencies publish guidelines on the use of AI in writing and reviewing grant proposals. We comply with these guidelines in their currently valid form.

We may use AI to generate or refine ideas, provided this is done in compliance with data protection regulations and in accordance with the rules for ensuring good scientific practice and preventing scientific misconduct. We also take research security issues into account: We are aware that the data entered is reused by AI tools. This can lead to difficulties if we thereby violate export control [DE] regulations.

Furthermore, AI can be used to refine texts, provided the same rules are followed and the content is appropriately labeled (see Academic Writing).