The editorial process at our leading information systems journals has been pivotal in shaping and growing our field. But this process has grown long in the tooth and is increasingly frustrating and challenging its various stakeholders: editors, reviewers, and authors. The sudden and explosive spread of AI tools, including advances in language models, make them a tempting fit in our efforts to ease and advance the editorial process. But we must carefully consider how the goals and methods of AI tools fit with the core purpose of the editorial process. We present a thought experiment exploring the implications of two distinct futures for the information systems powering today’s journal editorial process: an AI-augmented and an AI-driven one. The AI-augmented scenario envisions systems providing algorithmic predictions and recommendations to enhance human decision-making, offering enhanced efficiency while maintaining human judgment and accountability. However, it also requires debate over algorithm transparency, appropriate machine learning methods, and data privacy and security. The AI-driven scenario, meanwhile, imagines a fully autonomous and iterative AI. While potentially even more efficient, this future risks failing to align with academic values and norms, perpetuating data biases, and neglecting the important social bonds and community practices embedded in and strengthened by the human-led editorial process. We consider and contrast the two scenarios in terms of their usefulness and dangers to authors, reviewers, editors, and publishers. We conclude by cautioning against the lure of an AI-driven, metric-focused approach, advocating instead for a future where AI serves as a tool to augment human capacity and strengthen the quality of academic discourse. But more broadly, this thought experiment allows us to distill what the editorial process is about: the building of a premier research community instead of chasing metrics and efficiency. It is up to us to guard these values.