Learn / AI Security Engineer
AI Security Engineer Track

Prompt injection, jailbreaks, agent security, and AI compliance — the newest engineering role on the market.

The five stages

  1. 1

    Foundation

    In progress

    What AI security engineers do, why AI systems have unique attack surfaces, prompt injection fundamentals. Four concept pages, no code.

    8 items1.5 hrs
    0 of 8 completed

  2. Core Skills

    Submit boss to unlock

    Jailbreak taxonomy, defence in depth, red-teaming methodology, adversarial examples. Concept pages plus problems.

    8 items2.5 hrs
    Submit the Stage 1 boss to unlock

  3. Applied Practice

    Submit boss to unlock

    RAG system security, agent exfiltration attacks, privacy in ML, AI compliance. Medium problems with real attack scenarios.

    6 items3.5 hrs
    Submit the Stage 2 boss to unlock

  4. Interview Ready

    Submit boss to unlock

    Federated learning security, backdoor detection, watermarking and provenance. Hard problems with golden-answer comparison.

    5 items4.5 hrs
    Submit the Stage 3 boss to unlock

  5. Final Prep + Certificate

    Pro

    Company-tagged problems only. Required 15-minute simulator capstone. Completion generates a shareable certificate.

    3 items3.5 hrs
    Unlock with Pro