1. Basic principles

The rapid developments in the field of artificial intelligence (AI) promise a wide range of benefits for society in general. This also applies to research and scholarly publishing. However, these developments also harbor a number of risks. With this AI guideline, Berlin Universities Publishing (BerlinUP) addresses these risks in relation to scholarly publishing and regulates the handling of artificial intelligence and the use of AI tools for BerlinUP publications.

As orientation, the guideline is based on the Ethics Guidelines for Trustworthy AI [1] mandated by the European Commission, the position paper of the Committee of Publication Ethics (COPE) Authorship and AI [2], the recommendations of the World Association of Medical Editors (WAME) Chatbots and Generative Artificial Intelligence in Relation to Scholarly Publications [3] and the Statement by the Executive Committee of the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) on the Influence of Generative Models of Text and Image Creation on Science and the Humanities and on the DFG’s Funding Activities [4].

Artificial intelligence is an umbrella term for applications in which machines perform human-like intelligence. In the field of scholarly publishing, generative AI tools that create new content in the form of text, images, audio, video, software code or data sets using natural language input ("prompts") are particularly relevant.

Generative AI tools are based on machine learning technology, in which patterns are recognized autonomously on the basis of training data and algorithms are trained to match the respective input. Well-known examples include text-generating chatbots such as ChatGPT, Copilot, Gemini or LLaMA as well as image-generating AI tools such as DALL-E, Stable Diffusion or Midjourney.

A trustworthy AI is characterized by the fact that it is lawful, ethical and robust. The most important ethical principles in the use of AI tools include respect for human autonomy, prevention of harm, fairness and explicabilty. BerlinUP derives the following rules of practice for dealing with AI in scholarly publishing from these principles, which are regularly evaluated due to dynamic developments.

2. Rules of practice

AI tools cannot assume authorship: According to COPE, WAME and DFG, AI tools cannot fulfill the requirements of authorship or intellectual property because they cannot take responsibility for the submitted work. AI tools are neither natural persons nor legal entities and therefore cannot regulate any transfer of rights or license agreements and cannot identify any possible conflicts of interest. BerlinUP agrees with this view and does not accept submissions in which AI tools are listed as authors or co-authors.

The use of AI tools must be described transparently: The use of AI tools in the creation of content in a submitted work (e.g. texts, tables, diagrams, images, audio, videos, software codes, data sets) must be described transparently. It must be clear which content was generated using AI tools, which AI tools were used and how the AI tools were used. This also applies to content for internal communication that is not intended for publication, such as abstracts or exposés, as well as to the use of AI tools in the production process for books and journals.

See also: Handout for Citing AI Tools

Authors are responsible for their submitted work: The authors are fully responsible for the content of their submitted work, and thus also for the parts that were created using AI tools. This includes the accuracy of the presentation, the omission of plagiarism and the appropriate citation of all sources. Authors are therefore liable for violations of good research practice and the ethical principles of scholarly publishing.

---

[1] High-Level Expert Group on Artificial Intelligence (2019): Ethics Guidelines for Trustworthy AI. https://digital-strategy.ec.europa.eu/en/library/ethics-guidelines-trustworthy-ai

[2] COPE (2023): Authorship and AI Tools. https://publicationethics.org/cope-position-statements/ai-author

[3] WAME (2023): Chatbots and Generative Artificial Intelligence in Relation to Scholarly Publications. https://doi.org/10.1080/03007995.2023.2286102

[4] DFG (2023): Statement by the Executive Committee of the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) on the Influence of Generative Models of Text and Image Creation on Science and the Humanities and on the DFG’s Funding Activities. https://www.dfg.de/resource/blob/289676/89c03e7a7a8a024093602995974832f9/230921-statement-executive-committee-ki-ai-data.pdf