AI Safety Red-Teamer — English & Kannada Bilingual Specialist

About the role

We're building a global red-teaming bench to help make AI systems safer, more accurate, and more culturally aware. We're looking for native-level English and Kannada speakers to join a distributed team of human data experts who stress-test AI models, surface weaknesses, and produce high-quality adversarial data used to train safer, more robust AI systems.

This is sensitive, high-impact work: you'll be reviewing and probing model outputs that touch on topics like bias, misinformation, harmful content, and edge-case reasoning failures. Your judgment and linguistic precision directly shape how AI systems behave for millions of Kannada and English speakers worldwide.

What you'll do

• Craft adversarial prompts and inputs designed to expose model vulnerabilities, biases, and failure modes
• Evaluate AI-generated responses in both English and Kannada for accuracy, safety, tone, and cultural nuance
• Document and categorize harmful, misleading, or low-quality outputs with clear rationale
• Translate, localize, and validate red-team datasets to ensure linguistic and contextual fidelity
• Collaborate with a distributed team of reviewers and safety researchers to refine evaluation guidelines
• Provide structured feedback that helps improve model alignment and reduce real-world harm
• Maintain strict confidentiality and handle sensitive content professionally and objectively

Requirements

• Native or near-native fluency in both English and Kannada (reading, writing, and nuanced comprehension)
• Strong analytical and critical thinking skills, with the ability to spot subtle biases, inconsistencies, or safety risks in text
• Comfort engaging with sensitive or adversarial content in a professional, detached manner
• Excellent written communication skills for documenting findings clearly
• Reliable internet access and the ability to work independently on a remote, flexible schedule
• Prior experience with content moderation, translation, linguistics, data annotation, or AI evaluation is a plus

Compensation

This role pays $20–$22 per hour, remote and flexible. Multiple slots are available for qualified bilingual candidates.