AI Red Team Engineer

• Conduct adversarial testing on LLM-powered systems to identify vulnerabilities such as jailbreaks, prompt injection, and data leakage.
• Create clear evidence and documentation for customer demos and security reviews.
• Write Python automation to run attacks, score responses, and generate repeatable test reports.
• Build and maintain an attack library of prompts, test cases, rubrics, and regression tests.

This role directly enhances the safety and security of AI systems, protecting users from malicious attacks and ensuring responsible AI deployment.

• Strong understanding of LLM vulnerabilities and adversarial testing techniques.
• Proficiency in Python for automation and scripting.
• Experience with security testing and documentation of findings.
• Ability to build and maintain test libraries and regression tests.

Compensation and benefits not specified in the posting.

White Circle is an organization focused on AI safety and security, working to identify and mitigate vulnerabilities in AI systems.