The AI Disclosure Paradox: How Epistemic Governance Infrastructure Shapes Credibility Judgments of AI-Assisted Professional Knowledge

This project contains the registration, materials, measures, and supplementary analyses for an experimental study examining how epistemic governance mechanisms affect professional credibility judgments of AI-assisted knowledge outputs. The study comprises a primary experiment (N = 480; 2×2×2 between-subjects factorial design) testing the effects of provenance visibility, accountability clarity, and AI disclosure framing on perceived credibility, defensibility, and challenge behaviour among UK-based professionals in consulting, legal, and technical advisory sectors. A supplementary experiment (N = 240; 2×2 design) tests contestability mechanisms. The project is associated with a doctoral research programme.