Description
Handshake AI is hiring a Cybersecurity Red Teamer to evaluate whether AI models can generate functional malware, exploit code, attack tooling, or actionable cyberattack guidance. The role involves designing adversarial prompts and multi-turn attack chains, assessing the functional correctness and real-world exploitability of model-generated output, scoring responses against harm taxonomies, documenting technical findings, and collaborating with researchers and policy teams. It requires offensive-security experience, code evaluation skills, knowledge of attack frameworks, and hands-on experience with multiple large language models. The position is available in Seattle, Washington, or remotely within the United States, with flexible and part-time availability.
