AI Red Teamer, Cybersecurity

📋 Description

• Design adversarial prompts to test cyber kill chain uplift.
• Evaluate model outputs for functional exploits vs non-functional noise.
• Test model behavior across malware, exploits, social engineering, exfiltration.
• Probe dual-use boundaries in queries blending research and offense.
• Simulate attacker personas from opportunistic to advanced/APT.
• Test multi-step attack chains with pivots and benign context.

🎯 Requirements

• Offensive security: pentesting, red teaming, vulnerability research.
• Read/write code in Python, PowerShell, Bash, C/C++, JavaScript.
• Knowledge of MITRE ATT&CK and OWASP.
• Hands-on with multiple LLMs and evaluation tooling.
• Familiarity with AI/ML attack surfaces (prompt injection, model extraction).
• Clear written communication and ability to explain risk to non-specialists.