Red Teaming | Generative AI Analyst - California

What You’ll Do

• Interact with generative AI models using project-provided guidelines, safety taxonomies, and attack-vector guidance.
• Create and evaluate prompts designed to test model behavior across safety-related categories.
• Identify where model responses become unsafe, noncompliant, inconsistent, or otherwise problematic.
• Document model breakability, effort level, point of failure, and relevant category alignment.
• Review text, image, audio, video, or other multimodal content as required by the workflow.
• Apply detailed guidelines consistently across short, high-volume production sprints.
• Use sound judgment to evaluate ambiguous, edge-case, or policy-sensitive outputs.
• Conduct self-review to ensure work is accurate, complete, and aligned with project expectations.
• Flag unclear guidelines, tooling issues, or recurring model behavior patterns.
• Participate in calibration, feedback, and quality review sessions to improve consistency.
• Maintain readiness to pivot quickly between different red teaming runs when active work is launched.

Requirements:

• Native-level or near-native English proficiency with excellent written communication skills.
• Work Authorization is required for the role.
• Strong creative writing ability and comfort constructing varied prompts.
• Experience with red teaming, safety data annotation, content evaluation, safety review, content moderation, QA, or AI model evaluation preferred.
• Strong attention to detail and ability to follow complex project guidelines.
• Ability to think critically and evaluate open-ended model responses.
• Comfort working with sensitive, adult, NSFW, or policy-relevant content where required.
• Interest in generative AI, AI safety, large language models, or emerging AI technologies.
• Ability to work quickly and accurately during short production windows.
• Bachelor’s degree or equivalent practical experience preferred.

Ways to Stand Out from the Crowd

• Background in creative writing, English, linguistics, journalism, communications, policy, trust and safety, or content moderation.
• Experience evaluating generative AI prompts and responses.
• Familiarity with AI safety, red teaming, jailbreak testing, RLHF, or model evaluation workflows.
• Experience working with safety taxonomies, policy guidelines, evaluation rubrics, or defect categories.
• Prior experience reviewing sensitive, adult, NSFW, or policy-relevant content in a professional setting.
• Experience with multimodal AI workflows involving text, image, audio, or video.
• QA/testing experience within AI, data operations, content review, or annotation environments.
• Ability to explain a repeatable approach for staying consistent during high-volume, judgment-based work.