Summary: Researchers are experimenting with generative AI tools like ChatGPT to assist in writing research papers, but concerns about the potential rise of poor-quality or error-strewn manuscripts threaten scientific communication. Publishers are worried about the accessibility of generative AI making it easier to create fake papers. Some journals are addressing this by banning LLMs altogether or insisting on transparency in their use. Additionally, concerns about equity and bias in generative AI use in science are being raised, particularly as non-native English speakers may benefit from such tools. Generative AI is also changing how research is published and disseminated, with more interactive formats emerging.