AAAI: Adversarial AI-Assisted Inquiry: An Experimental Proposal

We propose Adversarial AI-Assisted Inquiry (AAAI) as an experimental methodology for generating critical perspectives during argument development. Unlike conventional uses of AI as research assistants, AAAI employs AI models to simulate hostile examination of arguments before peer review. This paper articulates the theoretical motivation, describes an initial protocol, acknowledges significant limitations, and proposes empirical research needed to evaluate whether AAAI offers genuine methodological value. We emphasize that AAAI is a tool requiring careful human judgment, not a replacement for traditional peer review or a certification of intellectual rigor.