Generative Artificial Intelligence could strengthen open science by improving accessibility, findability, and usability. Still, under less favourable conditions, it could also weaken scientific exchange by obscuring provenance, deepening dependency on opaque infrastructures, and reinforcing inequalities and increasing synthetic overload, in which large volumes of AI-generated scientific content overwhelm the capacity of researchers and systems to verify, interpret, and trust information. This report describes how Generative AI is reshaping scientific exchange within the broader framework of open science. It shows that the main policy challenge is not whether Generative AI will be used in scientific exchange, but under what institutional, infrastructural, and normative conditions it can support reliable research, and knowledge valorisation. The European Union should treat scientific exchange as public-interest infrastructure and enact policies to support skills, strengthened provenance, transparency, reproducibility, inclusion, and strategic capacity in an AI-mediated research environment. This report provides specific policy options tackling the risks that can be associated with the use of AI in scientific exchange.
