AI-Augmented Research Without Replacing Evidence: A Methodological Framework for MIND Plus

This working paper presents the methodological principles of MIND Plus, an independent research and publishing project that uses artificial intelligence to expand the scale of scholarly discovery, comparison, synthesis and question generation without delegating scientific authority to AI systems. The framework treats generative AI as an analytical instrument operating inside a traceable evidence architecture. Claims are expected to remain connected to identifiable sources; disagreement and uncertainty are preserved rather than edited away; and final responsibility for research direction, interpretation and publication remains human. The paper describes a workflow that combines open scholarly infrastructures, citation mapping, structured comparison, verification and critical analysis. It also defines limits for AI use, including the risks of hallucination, automation bias, conceptual flattening and false comparability. The central proposal is that AI can increase the breadth and speed of research only when amplification is paired with explicit epistemic controls. Independence statement: MIND Plus is an independent research and publishing project. The project is not operated by, and this paper does not represent the institutional positions of, SEEDUC-RJ or any other employer or educational institution. AI-use disclosure: Artificial intelligence was used for structuring, drafting assistance, bibliographic discovery and editorial revision. The author directed the research question, approved the final argument and remains responsible for the content.