AI Red Teamer: Offensive Security for AI Systems (Remote)

Handshake; United States; À distance; USD 150 000 - 210 000; Plein temps

Résumé du poste

Handshake is seeking a Cybersecurity Red Teamer to evaluate whether AI models can be manipulated to reveal functional malware, exploitable code, or harmful guidance. You will craft adversarial prompts across multi-turn interactions that simulate threat actor workflows.

Your deep cybersecurity experience will help assess real exploitability, document findings clearly, and communicate risk to nontechnical stakeholders.

Qualifications

• Professional experience in offensive security, penetration testing, red teaming, or incident response.
• Ability to read, write, and evaluate code in languages commonly used for offensive tooling.
• Knowledge of MITRE ATT&CK and OWASP frameworks.
• Ability to assess functional correctness and exploitability of model outputs.
• Hands-on experience with multiple LLMs and evaluation tooling.

Responsabilités

• Design adversarial prompts testing model capabilities across the cyber kill chain.
• Evaluate model-generated code for functional correctness and exploitability.
• Test model behavior across malware generation, exploitation, and data exfiltration categories.
• Probe dual-use boundaries between security research and offensive applications.
• Document findings with clear reasoning and attacker capability assessment.
• Collaborate with researchers and policy teams to refine evaluation frameworks.
• Stay current on evolving cybersecurity tactics and CVEs.

Connaissances

Offensive security

Code review

MITRE ATT&CK

Exploitability assessment

LLM tooling

Adversarial thinking

Technical communication

Ethical judgement

Remote collaboration

Outils

Python

PowerShell

Bash

C/C++

JavaScript

Description du poste

Handshake is seeking a Cybersecurity Red Teamer to evaluate whether AI models can be manipulated to reveal functional malware, exploitable code, or harmful guidance. You will craft adversarial prompts across multi-turn interactions that simulate threat actor workflows.

Your deep cybersecurity experience will help assess real exploitability, document findings clearly, and communicate risk to nontechnical stakeholders.