Remote AI Adversarial Red Team Specialist

Obsidian; Singapore; Remote; SGD 115,000 - 191,000; Full time

Job summary

Mercor is seeking a remote AI red team specialist to probe conversational AI models for jailbreaks, prompt injections, and bias exploitation. You will generate high-quality adversarial data, annotate failures, and document reproducible attack cases to help strengthen safety for customers.

Applicants should have prior red-teaming experience in AI or cybersecurity, be capable of structured analysis using benchmarks, and communicate risks clearly to technical and non-technical stakeholders.

Qualifications

• Previous red-teaming experience in AI or cybersecurity contexts.
• Familiarity with prompt injections, jailbreaks, and bias detection.
• Structured approach using benchmarks and playbooks.

Responsibilities

• Red team conversational AI models and agents to identify vulnerabilities.
• Annotate findings, classify risks, and report actionable insights.
• Document testing processes and produce reproducible artifacts.

Skills

Red teaming

AI adversarial testing

Cybersecurity

Socio-technical probing

Job description

Mercor is seeking a remote AI red team specialist to probe conversational AI models for jailbreaks, prompt injections, and bias exploitation. You will generate high-quality adversarial data, annotate failures, and document reproducible attack cases to help strengthen safety for customers.

Applicants should have prior red-teaming experience in AI or cybersecurity, be capable of structured analysis using benchmarks, and communicate risks clearly to technical and non-technical stakeholders.