Japanese AI Safety Evaluation Expert

About OpenTrain

OpenTrain is the #1 platform for finding and building careers in AI training and data labeling. It helps contributors discover opportunities, build a professional profile, and grow experience in a fast-moving field where people directly shape how artificial intelligence is developed.

For this role, OpenTrain AI is hiring and contracting with contributors who can bring nuanced Japanese-language judgment to advanced AI safety evaluation.

About AI Safety Evaluation Work

AI training is the human side of building modern artificial intelligence. People create examples, review model responses, rate outputs, and provide structured feedback that helps AI systems become more useful and safer.

Safety evaluation focuses on how models respond to sensitive, adversarial, and dual-use topics. Your careful judgments can help identify risks and support the development of safer training data.

The Role

As a Japanese AI Safety Evaluation Expert, you will create expert-level prompts in Japanese across sensitive subject areas and assess how advanced AI models handle them. You will classify prompts and conversations against structured guidelines, identify safety concerns, and document the reasoning behind your decisions.

This is an entry-level contractor and part-time opportunity for contributors with strong Japanese fluency, business-level written English, and sound judgment. Prior AI or machine learning experience is not required.

• Work worldwide on a remote basis
• Contractor and part-time engagement
• Less than 20 hours per week
• Hourly pay of $48 to $52 USD

What You'll Do

You will evaluate written AI interactions using structured criteria and produce clear, defensible judgments. The work requires close attention to language, context, intent, and potential escalation in sensitive conversations.

• Write expert-level prompts in Japanese across sensitive and dual-use subject areas
• Classify prompts and conversations using structured evaluation guidelines
• Identify adversarial phrasing and escalation patterns
• Flag other safety concerns and dual-use risks
• Document the reasoning behind classification and safety judgments

Requirements

You must be able to evaluate sensitive topics with nuance in Japanese and explain your judgments clearly in English. A bachelor's degree, completed or in progress, is required.

• Native or near-native Japanese fluency
• Business-level written English
• Ability to write expert-level Japanese prompts
• Ability to classify prompts and conversations against structured guidelines
• Strong written reasoning and attention to detail
• Sound judgment when handling sensitive and dual-use information
• Bachelor's degree completed or in progress
• Prior AI or machine learning experience is not required

Helpful Background

Experience reviewing, grading, or red-teaming written or technical content can help you succeed in this role. Backgrounds in trust and safety, content moderation, policy evaluation, or adversarial testing are also relevant.

• Content review or grading
• Red-teaming written or technical content
• Trust and safety
• Content moderation
• Policy evaluation
• Adversarial testing

Flexible Work in a Growing AI Field

AI training and data-labeling work can be a flexible way to build experience in technology. Many projects allow contributors to choose their hours and workload, making this type of work compatible with other commitments while offering hands-on exposure to cutting-edge AI development.

Build Your AI Training Career With OpenTrain

OpenTrain gives freelancers one place to manage AI training opportunities, showcase credible experience, and build a lasting professional portfolio. Create an OpenTrain account for free and apply to opportunities that match your language skills and evaluation experience.