AI Red Teamer

Job Summary

Creative and ethical thinkers will find a full-time remote opportunity as an AI Red Teamer, responsible for stress-testing large language models by designing adversarial prompts that expose vulnerabilities and ensure AI safety across various risk categories.

Key responsibilities

• Craft creative prompts and multi-turn scenarios to stress-test AI guardrails across diverse risk categories
• Discover methods to bypass safety filters and evaluate model responses against structured harm taxonomies
• Document experiments and collaborate with engineers and researchers to share findings and strengthen defenses

Required qualifications

• Strong hands-on experience using multiple large language models (LLMs)
• Intuition for crafting adversarial prompts and familiarity with jailbreak or evasion techniques
• Creative problem-solving skills with a strong ethical judgment
• Self-directed and comfortable in feedback-heavy environments
• Experience or expertise in high-risk domains such as cybersecurity or medicine is a plus