Skip to content
Instant download · PDF + web edition · Free sample of every book · 30-day refund promise
Latent Library Browse books
Ethics & Governance

AI Security

Attacks, Defences and Secure Design for AI Systems

Pages
82
Chapters
66
Level
Intermediate–Advanced
Format
E-book

About this book

AI systems add new attack surfaces to familiar software: the data they learn from, the text they read and the actions they can take. This book explains the threats and the defences that work.

You will learn threat modelling for AI, prompt injection and jailbreaks, data poisoning and backdoors, adversarial examples, model and data extraction, privacy attacks, supply-chain risks in models and datasets, and the special risks of agents with tools. Defensive chapters cover secure architecture, least privilege, input and output controls, monitoring, red teaming and incident response.

The focus is defensive and practical: understanding attacks so you can design systems that resist them.

What you will learn

  • Threat modelling for AI systems
  • Prompt injection, jailbreaks and their limits as defences
  • Data poisoning, backdoors and adversarial examples
  • Model extraction, data leakage and privacy attacks
  • Supply-chain security for models and datasets
  • Secure architecture, monitoring, red teaming and incident response

Table of contents

  1. IntroductionIntroduction
  2. Chapter 1Why AI Security Is Different
  3. Chapter 2Security Principles That Still Apply
  4. Chapter 3Threat Modelling for AI Systems
  5. Chapter 4The AI Attack Surface
  6. Chapter 5Attacker Goals
  7. Chapter 6Risk Depends on Capability
  8. Chapter 7Security Requirements Up Front
  9. Chapter 8Shared Responsibility With Providers
  10. Chapter 9Prompt Injection
  11. Chapter 10Indirect Prompt Injection
  12. Chapter 11Why Prompt Injection Is Hard to Solve
  13. Chapter 12Defending Against Prompt Injection
  14. Chapter 13The Lethal Combination
  15. Chapter 14Jailbreaks
  16. Chapter 15System Prompt Leakage
  17. Chapter 16Sensitive Data Disclosure
  18. Chapter 17Securing Retrieval Systems
  19. Chapter 18Agents and Tool Use
  20. Chapter 19Insecure Output Handling
  21. Chapter 20Denial of Service and Cost Abuse
  22. Chapter 21Data Poisoning
  23. Chapter 22Backdoors
  24. Chapter 23Defending Training Pipelines
  25. Chapter 24Adversarial Examples
  26. Chapter 25Evasion of AI-Based Security Tools
  27. Chapter 26Model Extraction
  28. Chapter 27Training Data Extraction
  29. Chapter 28Membership Inference
  30. Chapter 29Model Inversion
  31. Chapter 30Differential Privacy
  32. Chapter 31Federated and Confidential Computing
  33. Chapter 32Watermarking and Fingerprinting Models
  34. Chapter 33The AI Supply Chain
  35. Chapter 34Malicious Model Files
  36. Chapter 35Dependency Management
  37. Chapter 36Model and Dataset Provenance
  38. Chapter 37Plugins and Integrations
  39. Chapter 38Infrastructure Security
  40. Chapter 39Secrets Management
  41. Chapter 40Identity and Access for AI Systems
  42. Chapter 41Multi-Tenant Isolation
  43. Chapter 42Secure Architecture Patterns
  44. Chapter 43Input Controls
  45. Chapter 44Output Controls
  46. Chapter 45Human Oversight
  47. Chapter 46Logging and Monitoring
  48. Chapter 47Red Teaming
  49. Chapter 48Security Evaluation Suites
  50. Chapter 49Incident Response for AI
  51. Chapter 50Kill Switches
  52. Chapter 51Governance and Policy
  53. Chapter 52Standards and Frameworks
  54. Chapter 53Training People
  55. Chapter 54AI-Enabled Attacks
  56. Chapter 55Using AI for Defence
  57. Chapter 56A Threat Model, Worked Through
  58. Chapter 57Secure Prompt Design
  59. Chapter 58Testing Agents in Sandboxes
  60. Chapter 59Securing Fine-Tuning and Training
  61. Chapter 60Images, Audio and Documents as Inputs
  62. Chapter 61Protecting AI Logs
  63. Chapter 62Measuring Security Posture
  64. Chapter 63Case Study: The Email Assistant
  65. Chapter 64Case Study: The Leaky Knowledge Base
  66. Chapter 65Case Study: The Poisoned Feedback Loop
  67. Chapter 66A Security Checklist
  68. GlossaryGlossary

Who this book is for

Security engineers, developers and architects who build or protect AI systems.

Details

FormatPDF + web edition
Length82 pages (approx.)
Chapters66
LevelIntermediate to Advanced
LanguageEnglish
TopicEthics & Governance
AuthorLatent Library Editorial

Keep reading

You may also like

All books
$14.99
Details