AI Security
A Practitioner’s Field Guide: the complete three-volume set
Each volume stands alone, with its own index. Read in order — Learn, Run, Operate — or take the one you need.
Securing AI and ML systems is the security discipline you already run (asset inventory, threat modelling, vendor due diligence, incident response), applied to systems with AI components. The three volumes take you from the full threat landscape, through running the programme, to the templates that make it operational.
Companion to the AI Security Assessment toolkit — the assessment and ready-to-use templates that put this book to work.
From the blog: Before your team pastes customer data into a chatbot — an essay-length preview of the same ground this set covers, free.
Foundations and the Threat Landscape
Volume 1 of the three-volume AI Security set: the foundations and the full attack surface, covering what AI security actually is, AI systems as security objects, the vulnerability taxonomy, threat modelling, the maturity spectrum, and the attack chapters: prompt injection, jailbreaks, training-data poisoning, model extraction and inversion, embedding-space attacks, the AI supply chain, and agentic systems. Volume 2 covers running the programme; Volume 3 is the Operator's Workbook. Each volume stands alone. Each is modestly priced in ebook.
Running the Programme
Volume 2 of the three-volume AI Security set: running the programme, covering AI vendor due diligence, AI incident response, red-teaming, tabletop scenarios, defensive AI in security operations, governance under NIST AI RMF and ISO 42001, the regulatory horizon (EU AI Act, GDPR), and maturity and cadence. Volume 1 covers foundations and the threat landscape; Volume 3 is the Operator's Workbook. Each volume stands alone. Each is modestly priced in ebook.
The Operator's Workbook
Volume 3 of the three-volume AI Security set: the Operator's Workbook. It holds the vendor due-diligence set, threat-modelling templates, incident response runbooks, the red-team coordination pack, governance templates, tabletop exercises, regulator engagement templates, extended case studies, and programme health, plus the template finder, per-tier implementation checklists, the AI Security Glossary, and the Maturity Rubric. Each volume stands alone. Each is modestly priced in ebook.
The set at a glance
Swipe sideways to see the full table →
| Volume | Pages | Depth | ISBN | Price |
|---|---|---|---|---|
| Volume 1 | 324 | 17 chapters | TBD | TBD |
| Volume 2 | 327 | 15 chapters | TBD | TBD |
| Volume 3 (Workbook) | 362 | 11 chapters | TBD | TBD |
Where to buy
AI Security will be available in paperback, hardcover, and ebook through print-on-demand retailers, each volume sold separately. The links activate the moment the titles list; until then, email us and we’ll tell you at launch.
Cover art, pricing, and availability are not yet final and will be updated when these titles are listed for sale. Books are sold through print-on-demand retailers (KDP, IngramSpark, Amazon, and independent bookstores) — not through the Sylvan Assurance toolkit store.