AI Peer Review: Founding Position Paper

AI-assisted peer review is not merely a response to a capacity crisis — it is an opportunity to make the review process better for everyone: faster for reviewers, fairer for authors, and more complete for the mathematical community as a whole. This founding position paper identifies the structural mismatch between AI-accelerated mathematical discovery and traditional peer review infrastructure, argues that AI-assisted review governed by explicit epistemic accountability frameworks is a win-win-win structural upgrade, and proposes a concrete four-stage hybrid architecture (AI triage → AI advocate-critic → human editorial decision → mixed re-review). An appendix situates the current moment within a 1980–2050 trajectory using MLB's 2026 Automated Ball-Strike system as a structural analogy. The lesson: done right, it makes the game better. Part of the AI Epistemic Governance Stack (Papers 1–10 at Zenodo). This is Paper 11. Related frameworks:- AEAL: 10.5281/zenodo.19565086- ZTEK: 10.5281/zenodo.19591564- AI Behavior Science Founding Paper: 10.5281/zenodo.19562751- Change Management Framework: 10.5281/zenodo.19591685