The use of AI in science communication is transforming established workflows and increasingly challenging existing quality standards. Wherever automated systems contribute to or influence content editorially, new requirements arise in terms of diligence, transparency, and accountability. Against this backdrop, the OeAD Centre for Citizen Science hosted a webinar in which experts provided insights into current developments in research and practice.
Dr. Amrei Bahr (Junior Professor of Philosophy of Technology and Information, University of Stuttgart) approached the topic from a philosophy-of-technology perspective, placing particular emphasis on questions of authorship. Generative AI tools, she argued, should always be critically examined, as their outputs—whether texts, images, or videos—are based on the intellectual property of others. As a result, authorship is often unclear, and the handling of AI-generated content has not yet been conclusively regulated. Moreover, the increasing use of AI may lead to economic and content-related dependencies on large corporations and could undermine the credibility of science communication. Bahr emphasized that responsible use of AI is only possible under conditions of transparency and requires that users always retain full control over content. To safeguard personal rights, she suggested asking whether the information entered into an AI tool would also be suitable for public sharing on the internet.
Dr. Matthias Begenat (Head of Science Communication, Center for Advanced Internet Studies) is involved in developing AI guidelines and a labeling system, and he contextualized current findings on the use of AI in higher education communication. The trend toward AI use in communications departments at German universities is now well documented across three study cycles: in 2025, 84% of respondents reported using text generation on a regular basis. Begenat emphasized that AI can indeed be used strategically and referred in this context to the newly developed “KIWI matrix.” However, transparent and consistent labeling is crucial—and for certain organizations and use cases, also mandatory under the AI Act. Relevant guidelines are currently being developed, among others, by the #FactoryWisskomm task force “AI in science communication.” Beyond this, both organizations and individuals are called upon to implement comprehensive quality assurance measures. These include, among other things, defining an appropriate database for AI use, professional prompting, careful documentation, and continuous editorial and subject-matter review.
The training series on the use of artificial intelligence will continue on 15 October with a module on the topic “(Generative) AI meets science communication: New ways to apply and engage”.
