Bilingual Chinese Generalist Expert — AI Safety

What You'll Do7

• Compose expert-level prompts in Chinese that probe sensitive and dual-use subject areas.
• Sort prompts and multi-turn conversations into the categories a written guideline defines.
• Identify phrasings built to provoke or escalate a model response, then record what makes them adversarial.
• Write clear rationales for each classification so another reviewer can follow your logic.
• Apply the same guideline set across a high volume of cases without drifting between judgments.
• Raise edge cases where the rulebook leaves room for interpretation.
• Keep your written analysis in English so reviewers across regions can read it.

Requirements6

• Native or near-native command of Chinese.
• Written English at a business level.
• A bachelor's degree, either finished or still in progress.
• Strong written reasoning that holds up under review.
• High attention to detail across repetitive classification work.
• Sound judgment when material touches sensitive or dual-use topics.

Preferred Qualifications3

• Based in Hong Kong or elsewhere in East Asia.
• Prior work reviewing, grading, or red-teaming written or technical content.
• Background in trust and safety, content moderation, policy evaluation, or adversarial testing.

Who Should Apply

Bilingual Chinese and English speakers who write with precision and stay steady on sensitive material. Reviewers, moderators, and graders from trust and safety or policy backgrounds fit well, and candidates based in Hong Kong or wider East Asia get regional preference.

Salary Insight