Advanced~8 hours6 lessons

LLM Security & Production Hardening

Secure AI systems against real threats: prompt injection, data exfiltration, PII leakage, and adversarial inputs. Build governance controls and compliance-ready architectures.

What you'll build: a secure input/output moderation proxy

Start Course

Curriculum

Module A: Attack Vectors

3 lessons

Lesson 1: AI Threat Modeling

Apply the STRIDE threat model to AI systems. Enumerate the unique attack surface of LLM-powered applications: the model, the prompt, the tools, the data pipeline, and the users.

55 min

Lesson 2: Prompt Injection — Defense in Depth

Understand direct and indirect prompt injection attack patterns. Build a multi-layer defense: input sanitization, output validation, privilege separation, and automated attack testing.

75 min

Lesson 3: PII Detection & Data Privacy in LLM Pipelines

Detect and redact personally identifiable information before it enters the LLM context. Implement anonymization, synthetic data generation, and differential privacy techniques for AI workflows.

70 min

Recommended Watch

Prompt Injection Attacks Explained

How prompt injection and jailbreak attacks work, and why they're the #1 LLM security threat.

Computerphile16:08

Module B: Governance & Compliance

3 lessons

Lesson 4: Secure Multi-Tenant AI Systems

Prevent data leakage between tenants in shared AI infrastructure. Implement context isolation, per-tenant rate limiting, audit logging, and data residency controls.

65 min

Lesson 5: AI Compliance — GDPR, HIPAA & the EU AI Act

Map AI system components to regulatory requirements. Understand right-to-erasure implications for vector databases, HIPAA constraints on clinical AI, and the EU AI Act risk tiers.

60 min

Lesson 6: Capstone — Harden a Production AI App

Take a vulnerable AI application, perform a threat model assessment, implement prompt injection defenses, add PII scrubbing, build an audit log, and validate hardening with an automated red-team test suite.

120 min

Recommended Watch

AI Governance & Responsible AI

Frameworks for AI governance — compliance, audit trails, multi-tenant isolation, and PII handling.

IBM Technology9:52

Course Quiz

5 questions · ~5 min · Pass 4/5 to unlock your badge

Start Quiz →