AI Safety Red Team Expert English Indonesian

About OpenTrain

OpenTrain AI is the hiring and contracting organization for this role and the #1 platform for finding and building careers in AI training and data labeling. It helps contributors discover specialized projects, build a portable AI training profile, and apply in minutes.

You will work remotely on flexible AI safety projects while contributing to a growing field that helps make advanced AI systems more robust, safe, and trustworthy.

• Remote contractor opportunity
• Part-time schedule of 20+ hours per week
• Compensation of $17 to $25 USD per hour
• Worldwide opportunity

About AI Safety Training

AI training is the human side of building artificial intelligence. People evaluate model behavior, prepare high-quality examples, identify failures, and provide structured feedback that helps AI systems perform more reliably.

Red teaming is a specialized form of AI evaluation. By probing models with adversarial inputs and documenting what goes wrong, contributors help identify vulnerabilities and risks before they affect real users.

• Work directly with conversational AI models and agents
• Create evaluation data that supports safer AI development
• Use structured taxonomies, benchmarks, and playbooks
• Contribute to cutting-edge AI safety work

The Role

OpenTrain is seeking an AI Safety Red Team Expert to conduct text-based safety evaluations of conversational AI models and agents. You will use an adversarial mindset to push systems toward their breaking points, generate high-quality red team data, and document vulnerabilities clearly.

The role requires native or bilingual fluency in English and Indonesian, prior red teaming experience, and the ability to communicate technical and non-technical risks in a clear, reproducible way.

• Experience level: Expert
• Languages: English and Indonesian
• Data type: Text
• Labeling focus: Evaluation, rating, red teaming, and classification

What You'll Do

You will investigate how conversational AI models and agents respond to adversarial scenarios, then turn your findings into consistent, actionable data and documentation.

• Red team models and agents using jailbreaks, prompt injections, misuse cases, bias exploitation, and multi-turn manipulation
• Annotate failures and classify vulnerabilities
• Flag systemic risks revealed during testing
• Follow taxonomies, benchmarks, and playbooks to keep evaluations consistent
• Produce reproducible reports, datasets, and attack cases
• Explain identified risks clearly to technical and non-technical stakeholders

Required Qualifications

This expert-level role is designed for candidates who have already worked in red teaming or a closely related area. You should be comfortable probing complex systems, applying structured evaluation methods, and adapting your approach across different projects and customers.

• Prior red teaming experience in AI adversarial work, cybersecurity, or socio-technical probing
• Curiosity and an adversarial mindset focused on testing system limits
• A structured approach using frameworks or benchmarks
• Strong written and verbal communication skills
• Adaptability across different projects and customers
• Native or bilingual fluency in both English and Indonesian

Additional Experience That Helps

The following backgrounds are welcome and may help you approach adversarial testing from different angles. They are listed as nice-to-have experience rather than required qualifications.

• Adversarial machine learning, including jailbreak datasets, prompt injection, RLHF or DPO attacks, and model extraction
• Cybersecurity experience in penetration testing, exploit development, or reverse engineering
• Socio-technical risk work involving harassment or disinformation probing, abuse analysis, or conversational AI testing
• Creative probing experience in psychology, acting, or writing for unconventional adversarial thinking

Why This Work Matters

Every major AI system depends on human evaluation and feedback. By uncovering failure modes and documenting them carefully, you will help shape how conversational AI behaves and support the development of safer, more trustworthy systems.

This work also offers a direct way to build experience in a fast-growing area of technology while contributing to projects that influence the future of AI.

• Play a direct role in improving AI safety and robustness
• Build specialized experience in AI red teaming
• Work remotely with flexible scheduling
• Contribute to a portable AI training career profile through OpenTrain

How to Apply

Create a free OpenTrain account and apply for this contractor opportunity in minutes. Be prepared to demonstrate your relevant red teaming experience, structured testing approach, communication skills, and English and Indonesian fluency.

• Apply through OpenTrain
• Contractor and part-time engagement
• Expected availability: 20+ hours per week
• Pay range: $17 to $25 USD per hour