Remote AI Red Team Specialist - Adversarial Testing

Obsidian; Singapore; Remote; SGD 153,000 - 230,000; Full time

Job summary

Mercor is seeking a remote red team specialist to probe conversational AI systems with adversarial inputs, jailbreaks, and prompt injections. You will generate high-quality human data, annotate failures, and surface vulnerabilities to help make AI safer for customers.

You’ll follow taxonomies and playbooks, document reproducible reports, and communicate risks clearly to technical and non-technical stakeholders.

Qualifications

• Prior red teaming experience in AI adversarial work or cybersecurity.
• Curious and adversarial, able to push systems to breaking points.
• Structured approach using frameworks or benchmarks.
• Excellent communication with technical and non-technical stakeholders.
• Adaptable and able to move across projects and customers.

Responsibilities

• Red team conversational AI models, jailbreaks and prompt injections.
• Generate high-quality human data by annotating failures and flagging risks.
• Apply taxonomies, benchmarks, and playbooks for consistent testing.
• Document reproducibly with reports and attack case datasets.

Skills

Red teaming

English

Indonesian

Cybersecurity

Structured approach

Job description

Mercor is seeking a remote red team specialist to probe conversational AI systems with adversarial inputs, jailbreaks, and prompt injections. You will generate high-quality human data, annotate failures, and surface vulnerabilities to help make AI safer for customers.

You’ll follow taxonomies and playbooks, document reproducible reports, and communicate risks clearly to technical and non-technical stakeholders.