Presentation Information
[U03-04]Responsible Use of AI Tools in the Transparent, Community-Driven, Not-for-Profit Publications of the European Geosciences Union (EGU)★Invited Papers
*Barbara Ervens1, Eduardo Queiroz Alves2 (1.CNRS, University Clermont Auvergne, France, 2.European Geosciences Union (EGU), Munich, Germany)
Keywords:
Scientific Communication,Publishing
As a community with a long-standing commitment to transparency, openness and rigorous scientific discourse, the publications of the EGU are well positioned to engage with challenges posed due to emerging AI technologies. EGU’s current guidelines explicitly permit assistive AI tools for language, grammar, and readability, helping to lower barriers for researchers who are non-native English speakers or who have limited access to professional editing support. At the same time, the guidelines require disclosure of any generative AI use and prohibit AI-generated text, interpretations and peer-review reports.
While compliance with these obligations cannot be fully verified by technical means alone, EGU’s open peer-review model, including the public discussion of preprints and the publication of all reports by reviewers, editors, authors, and community members, provides a powerful mechanism to identify and discourage inappropriate or fraudulent uses of generative AI. By keeping authors, editors, and referees accountable for scientific judgement and decision-making and by defining clear and uniform obligations across all roles, the guidelines promote fairness, inclusivity and trust while prioritizing human expertise and diverse perspectives at the core of the publication process.
In this presentation, I will outline the rationale behind EGU’s AI guidelines for scientific publications and how they are implemented within the interactive public peer-review process. I will also discuss open questions and new opportunities for responsible, community-driven use of AI in scholarly publishing.
While compliance with these obligations cannot be fully verified by technical means alone, EGU’s open peer-review model, including the public discussion of preprints and the publication of all reports by reviewers, editors, authors, and community members, provides a powerful mechanism to identify and discourage inappropriate or fraudulent uses of generative AI. By keeping authors, editors, and referees accountable for scientific judgement and decision-making and by defining clear and uniform obligations across all roles, the guidelines promote fairness, inclusivity and trust while prioritizing human expertise and diverse perspectives at the core of the publication process.
In this presentation, I will outline the rationale behind EGU’s AI guidelines for scientific publications and how they are implemented within the interactive public peer-review process. I will also discuss open questions and new opportunities for responsible, community-driven use of AI in scholarly publishing.