Publication Specifications
FORMATDRM-Free PDF
LENGTH35 Pages
TIERadvanced Edition
PUBLISHERBlizarra Ltd
Advanced35 pages • Instant Download
AI Safety, Guardrails & Red-Teaming
Defending against prompt injection, jailbreaks, data exfiltration, and model evasion attacks.
One-time purchase
£39.00
Instant Access
Encrypted payment processing via Stripe Checkout
DRM-free PDF e-book delivered immediately after payment
Purchase receipt emailed automatically
About this Book
If you connect an LLM to tools and sensitive data, it will be attacked. Learn the actual mechanics of indirect prompt injection, ASCII smuggling, recursive evaluation exploits, and data extraction. Implement dual-LLM arbiters, input sanitation, and programmatic safety guardrails.
Target Audience: Security engineers, compliance officers, and developers shipping customer-facing AI.
What You Will Learn & Build
Identify and neutralize direct and indirect prompt injection vectors
Defend against multi-step jailbreaks, roleplay bypasses, and encoding tricks
Deploy NeMo Guardrails and Llama Guard to enforce enterprise safety boundaries
Conduct automated red-teaming simulations and generate vulnerability reports
Table of Contents
1. Threat Modeling for LLM-Powered Applications (OWASP Top 10)Section 1
2. Indirect Prompt Injection via RAG & Web Browsing ToolsSection 2
3. Adversarial Jailbreak Techniques & Automated Red-TeamingSection 3
4. Guardrail Architectures: Input Screening, Output Filtering & Dual LLMsSection 4
5. Preventing Sensitive PII & API Key ExfiltrationSection 5
6. Incident Response & Threat Hunting in Model LogsSection 6
7. Automated Red-Team Pipelines & Continuous Adversarial Testing in CISection 7
8. Model & Supply-Chain RiskSection 8