Bilingual AI Safety Data Evaluator

The work

You will review AI-generated content in Spanish and English and assess whether the responses are safe, accurate, logical, and well explained. You will provide annotations and feedback that help improve AI model behavior.

The work can include sensitive or disturbing material. You will need to apply nuanced safety policies, spot risky responses, and stay consistent when reviewing difficult edge cases.

• Evaluate the quality of AI safety and reasoning decisions.
• Label and quality-check safety data in Spanish and English.
• Red-team AI responses by looking for edge cases and unsafe behavior.
• Rate generated text and provide clear feedback for model improvement.
• Apply safety policy judgments to identify risky or harmful responses.

What it pays and takes

This is a remote, hourly-paid contractor role. The work is available worldwide and requires strong Spanish and English skills.

• Pay: $14 to $24 USD per hour, with a listed rate of $20 per hour.
• Work arrangement: Remote and worldwide.
• Contract type: Contractor.
• Language: Spanish and English at C1+ level.
• Experience level: Intermediate.
• You should have strong judgment, careful attention to detail, and emotional resilience when reviewing sensitive material.

How it works

Apply on OpenTrain. The employer reviews applications there.

About AI training work

AI training work is the human work behind systems that generate and understand language. Reviewers rate model responses, identify problems, and provide feedback so AI can produce safer and more useful results.