Data Annotation Specialist, Safety

Responsibilities

• ▹Evaluate and improve model safety: label, rank, audit, and refine human- and model-generated text to improve safety, quality, and policy alignment
• ▹Apply nuanced safety judgment: assess model outputs against detailed safety guidelines and rubrics, including ambiguous and sensitive cases
• ▹Create prompts and safety test cases: write realistic user scenarios and adversarial examples to evaluate model behavior across safety categories
• ▹Support quality and calibration: identify annotation inconsistencies and provide feedback on recurring edge cases and model failures
• ▹Work with precision and independence: complete annotation tasks with strong attention to detail in a globally distributed, asynchronous team

Requirements

• ▹1+ years of experience in Content Moderation, Trust and Safety, AI data annotation, LLM evaluation, or a related analytical role
• ▹Experience applying detailed guidelines to complex and sensitive content, with strong contextual and sociocultural judgment
• ▹Emotional resilience: comfort working with unsafe, explicit, and/or toxic content, including sexual, violent, or psychologically disturbing material
• ▹Excellent command of written English and the ability to clearly justify content evaluations
• ▹Strong attention to detail and commitment to accuracy, with consistency across high-volume tasks
• ▹Strong execution in a remote, global, asynchronous team environment

Nice to have

• ▹Fluency in another language

Soft skills

What we offer

• ▹Fully remote, BYOD (bring-your-own-device) work within Canada or the United States
• ▹45 CAD/hour (Canada) or 40 USD/hour (US) contract rate, minimum 16 hours per week — includes hazard pay
• ▹12-month contract with potential for extension
• ▹Independent contractor status — may work with multiple clients simultaneously

About the company

Cohere's mission is to scale intelligence to serve humanity. It trains and deploys frontier models for developers and enterprises building AI systems for content generation, semantic search, RAG, and agents.