AI Red Teamer, Cybersecurity

Tasks

• Assess dual use boundaries in security research prompts
• Design adversarial prompts to test model cyber risk
• Document findings with technical reasoning
• Evaluate model generated code for functional correctness
• Refine cybersecurity evaluation frameworks and threat models
• Score model responses against harm taxonomies
• Simulate attacker personas to measure risk scaling
• Stay current on TTPs CVEs jailbreak techniques
• Test multi step multi turn attack chains
• Test offensive categories across cyber kill chain

Perks/Benefits

• N/A

Skills/Tech-stack

Adversarial examples | Bash | Binary Analysis | C# | C++ | Command and control | Command and control (C2) | Credential harvesting | Cyber Kill Chain | Data poisoning | Exploit Development | Incident Response | Java | JavaScript | Kill Chain | LLM | Malware analysis | Mitre Attack | Model extraction | OWASP | Penetration Testing | Powershell | Privilege escalation | Prompt injection | Python | Red Teaming | Reverse Engineering | Reverse Shell | Social engineering | Threat Intelligence | Training data | Training data poisoning | Vulnerability research