Hands-on LLM Security Labs

Live now

Hands-on LLM Security Labs

Intentionally vulnerable AI apps you exploit right in your browser: prompt injection, RAG data leaks, agent and tool abuse. Real targets, guided walkthroughs, and a certification track.

♦ Lab 01

Jackpot

Ten floors, ten broken AIs, one casino. Talk your way to the top using the real OWASP LLM Top 10: prompt injection, system prompt leakage, RAG poisoning, excessive agency. Simulated staff, not live models, so it is free and identical for everyone. No signup, runs in your browser.

Take the building →

More labs coming

You're on the list.

I'll email you once when the first lab goes live. No spam.

No spam. One email when the first lab drops.