THREAD SECURITYEDUCATION
AI SecurityIntermediate Level65 Hours Live

AI Security & LLM Red Teaming Masterclass

Exploit LLM pipelines, master indirect prompt injection, bypass NeMo guardrails, and mitigate agentic RCE threats.

Indirect Prompt Injection Vectors
NeMo Guardrails & Dual-LLM Auditing
VectorDB Data Poisoning Defense
65 Hours Practical Workload
1 Core Modules
1 Sandboxed Labs
Cryptographic TS-ID Verifiable

Course Overview & Objectives

The definitive cybersecurity curriculum for Generative AI. Audit foundation model integrations, execute direct and indirect prompt injections, circumvent output moderation rails, and defend production LLM pipelines from supply chain poisoning.

What You Will Master

  • Execute direct and indirect prompt injection attacks against multimodal foundation models
  • Test and bypass semantic guardrails built with NeMo Guardrails and Llama Guard
  • Audit autonomous AI agents for unsafe tool invocation and Server-Side Request Forgery (SSRF)
  • Implement dual-LLM verification architectures and input sanitization filters

Prerequisites

  • Basic Python programming
  • Understanding of LLM prompt engineering concepts

Platforms & Tools Covered

LangChainNeMo GuardrailsGarak LLM ScannerOpenAI APIChromaDB

Detailed Curriculum Modules

1 modules structured from foundational theory through complex adversarial execution.

65 Total Workload Hours
MODULE 01

Prompt Injection Taxonomy & Attack Surfaces

1 Lessons

Direct jailbreaks, token smuggling, base64 payloads, and indirect injection via third-party web content.

Anatomy of Indirect Prompt Injection in RAG Pipelines
50m

Hands-on Virtual Sandbox Labs

Zero local hardware dependencies. Provisioned in cloud containers via browser terminal.

LAB 01~50 mins

NeMo Guardrails Bypass Sandbox

Craft adversarial prompts that circumvent programmable Colang safety rules.

Skills Tested:NeMo, Prompt Injection

Faculty & Lead Instructor

Direct weekly instruction, live office hours, and code-review feedback.

AR

Ananya Roy

Thread Security Education

Head of AI Vulnerability Research

Pioneering researcher specializing in adversarial prompt injection and safety guardrails for production enterprise models.

Frequently Asked Questions

Everything you need to know about scheduling, cohort admissions, and lab access.

Are these attacks demonstrated on real LLMs?

Yes, labs utilize open-source models (Llama 3, Mistral) and API integrations.

Ready to Master AI Security & LLM Red Teaming Masterclass?

Join the upcoming cohort. Seats are limited to maintain a high faculty-to-student ratio and rigorous sandbox feedback.