AI Red Teaming Generalist

About the Role

GraceMark Solutions is seeking AI Red Teaming Generalists to train generative AI models and applications, ensuring safe, accurate, and responsible content. This is an entry-level role for strong writers and analytical thinkers interested in AI safety and red teaming. You will test AI model and application behavior with safety-sensitive topics and edge cases by designing prompts, evaluating outputs, and documenting evidence professionally.

Key Responsibilities

• Create original prompts that probe model and application behavior in complex or sensitive scenarios.
• Review, rank, and assess model and application outputs for accuracy, safety, and quality across text, image, video, audio, and agentic behaviors.
• Identify factual errors, misleading claims, and unsafe outputs; flag and document issues with clear, concise rationales.
• Communicate findings with professionalism and audience-appropriate detail, following writing and format standards.
• Collaborate with peers and supervisors to meet project quality bars and deadlines.

Requirements & Tech Stack

• Bachelor’s degree in a relevant field or equivalent writing-centric experience.
• Excellent command of English grammar and style; ability to write clearly, concisely, and professionally.
• Ability to work with sensitive or graphic material.
• Strong attention to detail; structured, evidence-based judgment using templates, rubrics, and scorecards.
• Reliable execution, time management, and responsiveness in a collaborative environment.

Preferred Backgrounds

• Writing-intensive fields (copywriting, journalism, editing, education).
• Experience in content moderation, gaming communities, business, and/or digital cultures.
• Interest in AI/ML technologies and their societal impact.

Work schedule: Hybrid - Tuesday-Thursday in-office (9 AM - 5:30 PM), Monday and Friday remote.