With the current rise of generative Artificial Intelligence (AI) tools, promising opportunities for the streamlining of access procedures and data querying are foreseeable. Conversational AI capabilities based on large language models have also the potential to facilitate large-scale genomic data-sharing from multiple available open-access datasets online. If this technology is to be rolled out in a responsible manner, however, guidance on the way these technologies can better account for attribution, privacy, bias, and accountability will be necessary. This working session will bring together members of the GA4GH community and the REWS Generative AI study group to brainstorm points to include in a guidance document that can promote the responsible use of conversational AI in the context of large-scale, open-access genomic databases.