
Episode #38
Agentic AI: Are You Ready for the Regulators?
As financial institutions put AI agents to work across fraud and AML, regulators will want to know more than what the technology can do. Can you explain its decisions, prove it was properly tested, reconstruct what happened, and show where human accountability begins and ends? We break down the governance questions CROs and CCOs need to address as agentic AI takes on more responsibility, including levels of autonomy, human oversight, explainability, audit trails, and what it takes to make AI-driven decisions defensible. Go deeper: The CRO & CCO’s Guide to Agentic AI https://www.datavisor.com/intelligence-center/ebooks/the-cro-cco-guide-to-agentic-ai What CROs & CCOs Need to Know About Agentic AI https://www.datavisor.com/blog/what-cro-cco-need-to-know-about-agentic-ai Chapters 00:00 The agentic AI governance challenge 02:45 Why traditional governance falls short 04:53 When AI risks compound 07:23 Five levels of AI autonomy 10:30 The explainability problem 11:09 How AI explanations fail 12:51 From signal to evidence 13:37 The Plain English Test 14:56 Building an exam-ready audit trail 16:30 Why human overrides matter 17:36 Governance by design 18:10 Knowledge Graphs and evidence 19:03 Governing a digital workforce 20:10 The risk of automation bias

