Skip to main content
Innodata Inc. logo

Red Teaming

Innodata Inc.
1 hour ago
Full-time
Remote
Taiwan

Red Teaming Prompt Writer - Mandarin


We’re hiring bilingual Mandarin & English-speaking Red Teaming Prompt Writers for a global, remote AI Safety & Adversarial Testing project.


You’ll test cutting-edge language and multimodal AI models by creating challenging prompts, identifying weaknesses, and uncovering potential safety risks before these models reach the public.

Role Details

  • Position: Red Teaming Prompt Writer
  • Languages: Mandarin Chinese & English
  • Work Mode: 100% Remote
  • Engagement: Freelance
  • Schedule: Flexible
  • Duration: 30 days
  • Compensation: $12 USD/hour

What You’ll Do


  • Create creative adversarial prompts, jailbreak scenarios, and edge-case tests for AI models
  • Test and evaluate AI responses across text, image, video, and audio
  • Identify safety vulnerabilities, unexpected behaviours, and model weaknesses
  • Assess whether model responses follow safety guidelines and instructions
  • Identify cultural, linguistic, and contextual nuances that may affect AI behaviour
  • Document testing intent, observations, and findings clearly
  • Adapt testing strategies as model behaviour and project guidelines evolve
  • Collaborate with QA and project teams to report findings


What We’re Looking For

  • Native or near-native Mandarin Chinese proficiency
  • Strong English communication and writing skills
  • Strong knowledge and understanding of Chinese culture, including social norms, customs, traditions, values, current cultural context, and linguistic nuances
  • Ability to understand the ins and outs of Chinese culture and recognize cultural references, sensitivities, implicit meanings, and context that may not be obvious to non-native speakers
  • Excellent understanding of Mandarin language, context, and cultural nuances
  • Strong analytical and critical-thinking skills
  • Creative approach to problem-solving
  • Ability to identify inconsistencies, loopholes, cultural sensitivities, and unexpected AI behaviour
  • Comfortable working with AI tools and following detailed guidelines


Previous experience in any of the following is a plus:

  • AI / LLM evaluation
  • Red teaming
  • Prompt writing
  • Trust & Safety
  • Content moderation
  • AI data annotation
  • Quality assurance
  • Adversarial testing


Previous AI experience is not mandatory if you have strong Mandarin and English skills, deep knowledge of Chinese culture, analytical thinking, and an interest in testing AI systems.


⚠️ Important Content Notice

This project may involve sensitive, disturbing, or NSFW content as part of AI safety and adversarial testing. Please consider this before applying.


📝 Assessment & Selection

Candidates will be required to complete an online assessment before onboarding.

  • 80 questions
  • 17 sections
  • 90 minutes
  • Passing score: 80%


Please ensure you have 90 uninterrupted minutes available to complete the assessment.

Candidates who successfully pass the assessment will be contacted regarding the next steps.


🚀 Interested?


If you are fluent in Mandarin and English, have a strong understanding of Chinese culture and cultural context, and enjoy creative problem-solving, we’d love to hear from you.