AI Defense and Induced Hallucination: Fighting AI Attacks with AI: A Practitioner's Field Guide
The attacker already put AI in the loop. Do the same, faster, on your own terrain.
400+ pages. 440 pages in print.
In 2025 a state-sponsored group used an AI agent to run most of an espionage campaign, with humans stepping in at only a handful of decision points. Production AI assistants were turned against the organizations that deployed them. The defenses most teams run were built for a slower adversary.
This practitioner's field guide shows security teams how to answer machine-speed attacks with machine-speed, under human control: threat models built on MITRE ATLAS and ATT&CK, an AI-augmented SOC with tiered autonomy and human gates, detection engineered for agent tempo, hardened agents, continuous autonomous validation, governance engineers will follow, and incident response for a machine adversary.
It also answers the question defenders keep asking: can we make the attacker's AI hallucinate? The deception chapter shows how to fill your own terrain with decoys an agent will believe, act on, and report back to its operators as fact.
Fourteen chapters in three parts (The New Offense, The New Defense, The Program), three appendices, sources cited throughout, and a reader's index. 440 pages. Written by a security operations lead who protects regulated medical imaging systems inside hospitals, where the systems also have to stay available for patient care.
Customer Reviews
ThriftBooks ® and the ThriftBooks ® logo are registered trademarks of Thrift Books Global, LLC
