AI Safety Red Teamer — English & Gujarati Bilingual Specialist

About the role

We're building a global red team of language and safety specialists who stress-test AI models before they reach the real world. This role is focused on adversarial probing, output review, and vulnerability discovery for AI systems operating in **English and Gujarati**. You'll be part of a distributed team that helps make AI safer, fairer, and more reliable for millions of end users by finding the cracks before anyone else does.

This is sensitive, high-impact work. You'll be reviewing AI-generated content that may touch on bias, misinformation, harmful instructions, or other risky outputs — and your job is to identify, document, and help remediate these issues with precision and professionalism.

What you'll do

• Craft adversarial prompts and inputs designed to surface weaknesses, biases, or unsafe behaviors in AI models
• Review and annotate AI-generated outputs in both English and Gujarati for safety, accuracy, and appropriateness
• Identify patterns of harmful, biased, or misleading model behavior and document them clearly for engineering and safety teams
• Translate and localize red-teaming test cases to ensure cultural and linguistic nuance is captured accurately
• Collaborate with a distributed team of safety experts to refine red-teaming methodologies and reporting standards
• Maintain strict confidentiality and handle sensitive content with professional judgment

Requirements

• Native or near-native fluency in **both English and Gujarati** — written and spoken
• Strong analytical and critical thinking skills, with attention to nuance, tone, and cultural context
• Comfort engaging with sensitive or potentially disturbing content in a professional, objective manner
• Excellent written communication skills for documenting findings clearly
• Reliable internet access and ability to work independently in a remote setting
• Prior experience with content moderation, linguistics, translation, or AI evaluation is a plus but not required

Compensation

Hourly compensation of **$20–$22/hour**, paid for verified work hours. Flexible, remote, project-based engagement with opportunities for extended collaboration based on performance.