As we increasingly recognize the importance of multilingual scholarly communication, translation-friendly writing is gaining more and more attention. While many applications of generative artificial intelligence (genAI) tools are being explored by professionals and laypersons alike, these tools’ capacity to facilitate translation-friendly writing has not yet been fully explored. <!–more–> In this paper, we present a pilot study of the use of ChatGPT to pre-edit scholarly journal abstracts in preparation for translation from English to French and Arabic, comparing with a human edit using the same guidelines. We use quantitative and qualitative methods to explore the edits made and their effect on the quality of the translations ChatGPT ultimately produces. The results may help to indicate future directions for research, including evaluations of the tool’s potential to assist scholars in sharing their work, and may help to guide efforts to implement ChatGPT effectively and responsibly in scholarly communication.