Internal audit is entering the most consequential transformation in its modern history.
A new generation of "agentic" AI systems can now pursue objectives, use tools, adapt to changing conditions, and execute complex work under human supervision. For U.S. banking audit functions, this changes more than technology. It changes how assurance itself is performed.
In Agentic Auditor, veteran banking audit leader and regulatory-law-trained practitioner David Bakinsky provides a practical framework for deploying AI across internal audit while independently governing the institution's own AI systems.
Drawing from real-world banking audit experience, the book shows how audit functions can evolve from periodic, sample-based testing toward continuous risk sensing, full-population testing, AI-assisted controls testing, and defensible governance over autonomous systems operating across credit, fraud, AML/KYC, compliance, and operations.
More importantly, it addresses the profession's defining challenge:
How can internal audit responsibly use AI while also providing independent assurance over it?
Designed for Chief Audit Executives, audit leaders, risk professionals, compliance officers, regulators, and banking practitioners, Agentic Auditor delivers practical guidance rather than theoretical AI commentary.
Inside, readers will find:
- A working playbook of ten AI audit agents across the audit lifecycle
- A practical first-ninety-days implementation roadmap
- Sample audit programs and testing approaches
- Governance frameworks aligned to emerging U.S. banking expectations
- Guidance for preserving independence, evidence quality, and professional skepticism in AI-assisted engagements
Written in clear, practical language for working audit professionals, Agentic Auditor is a field guide for leaders preparing their audit functions for the next era of banking oversight.
