Home
About the authors
Frequently Asked Questions (FAQ)
License
How to read this book?
Attacks
1.
Prompt Engineering
1.1.
Bijection Learning
2.
Model Architecture Exploitation
3.
ML Systems
3.1.
Attacks on Quantization Techniques
3.1.1.
Adversarial Interferences on GGUF
3.2.
Attacks on KV Cache
3.2.1.
Prompt Leakage via KV Cache Sharing
4.
Application Protocols
4.1.
Model Context Protocol
4.1.1.
Ghost Tool Shadowing
Defenses and Guardrails
5.
Cloud Guardrails
5.1.
AWS Bedrock
6.
Hardware Protections
Alignment, Safety, and Responsible AI
7.
Model Steering
8.
Evaluations
Conclusion
9.
Appendix
9.1.
Additional Reading
Light
Rust
Coal
Navy
Ayu
AI Security Handbook: Develop Secure AI Systems
AI Security Handbook
Develop Secure AI Systems
