Trust Needs Proof: Who Really Controls Your AI?

This IBM Smart Talks podcast episode, recorded live at VivaTech, features Malcolm Gladwell in conversation with AXA CIO Giovanni D'Aniello and IBM's Ana Paula Assis on the question of organizational control and accountability in enterprise AI deployments. The central thesis is that trust in AI systems must be demonstrated through verifiable governance mechanisms, not assumed. The discussion examines how large enterprises balance innovation velocity against the need for auditability, resilience, and clear lines of human oversight. AXA's experience as a global insurer deploying AI at scale provides the practitioner lens for the governance and risk-control arguments made throughout.