AI Defense and Induced Hallucination is a 442-page practitioner's field guide to defending against AI-driven cyberattacks, by Michael Romansky. It shows security teams how to answer machine-speed attacks with machine-speed defense: threat modeling on MITRE ATLAS and ATT&CK, an AI-augmented SOC, detection engineering at machine speed, and deception that makes an attacking AI hallucinate, waste its budget, and expose itself. Read the first two chapters free.
Buy the full book, $19.99Reads instantly in this page-turning reader with read-aloud, plus the PDF to keep. 442 pages.
Fourteen chapters across three parts: the new AI-enabled offense, the AI-augmented defense, and building the program.
SOC leads, incident responders, detection engineers, security architects and CISOs who need to defend AI systems and answer AI-driven attacks with AI defense, including induced-hallucination tools that turn an attacker's own agent against them.
Michael Romansky has spent about twenty-five years in IT and the last fifteen in security operations, including the protection of regulated medical imaging software. He writes from the defender's chair, not the vendor's.