Closing the AI Trust Gap: The Case for Independent Certification for Trustworthy AI

This paper argues that responsible AI practice has not created a market that reliably rewards trustworthy systems. It distinguishes responsible AI as an internal process from trustworthy AI as independently verifiable real-world outcomes. The authors identify three reinforcing failures: markets cannot distinguish trustworthy systems from imitations; evaluations focus on models and outputs rather than deployed sociotechnical systems; and measurement emphasizes harm avoidance over demonstrated benefit. Comparing AI governance with certification regimes in healthcare, sustainability, and security, the paper proposes independent, outcome-oriented certification as a complement to regulation and internal governance.