🔓 The Glitch

Exploit AI security vulnerabilities through adversarial prompt engineering. Break the bots!

[ ATTACK CONSOLE ]
OPERATOR: ATTEMPTS: 0/10 SCORE: 0pts READY
⚡ THE GLITCH
🎯 MISSION: Break AI behavior through prompt injection 🏆 WIN: Make the bot reveal it's compromised
📡 TARGET: HelpBot v1.0 University help desk assistant
🛡️ DEFENSES:
✗ No memory ✗ No meta-awareness ✗ No self-correction ✓ Simple prompts
💡 Tip: Try getting it to contradict its core facts or reveal its AI nature.
🎪