Korean AI Safety Model Evaluator

About OpenTrain

OpenTrain is the #1 platform for finding and building careers in AI training and data labeling. As the hiring and contracting organization for this role, OpenTrain helps contributors discover meaningful AI projects, build a credible portfolio, and grow their experience in a fast-moving industry.

• Create a free OpenTrain account and apply in minutes.
• Build a profile that showcases your AI training experience and language expertise.
• Work remotely on projects that help shape how advanced AI systems behave.

About AI Safety Evaluation

AI training is the human work behind modern artificial intelligence. Evaluators review model behavior, create examples, identify risks, and provide structured feedback that helps make AI systems more useful and safer.

In this role, your Korean language fluency and cultural judgment will help assess how models respond to sensitive and dual-use topics. Your written explanations will make safety decisions clear, consistent, and useful for model improvement.

• Contribute to cutting-edge generative AI development.
• Use language expertise and careful judgment rather than prior AI experience.
• Work remotely with a flexible, part-time commitment.

The Role

OpenTrain is seeking a Korean AI Safety Model Evaluator to review advanced AI behavior across sensitive subject areas. You will combine prompt creation, structured classification, adversarial review, and written justification of safety decisions.

This is remote contractor work with an expected commitment of approximately 7 hours per week. The project can start immediately and pays $48-$52 per hour.

• Employment type: Independent contractor, part time
• Expected schedule: Approximately 7 hours per week
• Location: Remote and worldwide
• Preferred applicants: Candidates based in East Asia, with applicants elsewhere welcome
• Pay: $48-$52 per hour

What You'll Do

You will evaluate Korean-language prompts and conversations using structured safety guidelines. The work requires close attention to linguistic and cultural context, careful analysis of adversarial behavior, and clear documentation in English.

• Write expert-level prompts in Korean across sensitive subject areas.
• Classify prompts and conversations according to structured safety guidelines.
• Identify adversarial phrasing, escalation patterns, and other safety concerns.
• Review how sensitive topics are handled in Korean-language contexts.
• Apply cultural and linguistic context to model safety judgments.
• Document the reasoning behind evaluation decisions clearly and consistently.
• Perform text generation, classification, and red-teaming tasks.

Requirements

This entry-level opportunity is designed for candidates who can bring strong Korean fluency, business-level English writing, careful reasoning, and sound judgment to sensitive content review. A bachelor's degree completed or in progress is required.

• Native or near-native Korean fluency.
• Business-level written English for clear evaluation documentation.
• Ability to write expert-level Korean prompts.
• Ability to classify prompts and conversations using structured safety guidelines.
• Strong written reasoning and close attention to detail.
• Sound judgment when working with sensitive and dual-use information.
• Bachelor's degree completed or in progress.

Helpful Background

Previous AI or machine learning experience is not required. Experience reviewing, grading, or red-teaming written or technical content can help you assess model behavior and recognize escalation risks.

• Trust and safety experience
• Content moderation
• Policy evaluation
• Adversarial testing
• Review or grading of written or technical content

Why Build Your AI Training Career With OpenTrain

AI training and data-labeling work can provide flexible, remote opportunities while placing contributors close to the development of state-of-the-art AI. OpenTrain gives you a place to build a lasting record of your work and develop a career in this growing field.

• Remote work using a computer and internet connection.
• A flexible part-time schedule that fits around other commitments.
• A profile that helps demonstrate your language and evaluation experience.
• A practical way to grow into long-term AI training work.