This job post has expired on October 26, 2025. It is likely that the position has already been filled.
AI Red-Teamer — Adversarial Testing at Mercor
posted 1 year agoAI Red-Teamer — Adversarial Testing | $54–111/hr | Remote: US/UK/CA
Join Mercor as an AI Red-Teamer and play a pivotal role in advancing the safety and reliability of cutting-edge AI systems. This position offers flexible hourly engagement for professionals passionate about probing AI models and uncovering vulnerabilities.
- Key Responsibilities:
- Conduct adversarial testing on AI models, including jailbreaks, prompt injections, and misuse scenarios.
- Generate and annotate high-quality data, classify vulnerabilities, and flag systemic risks.
- Produce reproducible reports and datasets to strengthen customer AI systems.
- Support diverse projects, ranging from LLM jailbreaks to socio-technical abuse testing.
- Ideal Candidate:
- Experience in AI red-teaming, cybersecurity, or a strong AI background.
- Curious, structured, and communicative—able to explain risks to varied audiences.
- Adaptable and eager to work across multiple projects and customers.
- Preferred Skills:
- Adversarial ML, cybersecurity, socio-technical risk analysis, or creative probing.
- Why Mercor?
- Directly impact AI safety and robustness.
- Collaborate with top experts and leading AI labs.
- Flexible remote work within US, UK, or Canada time zones.
Mercor is a premier talent marketplace connecting experts with frontier AI projects. Compensation varies by project scope and expertise required.
How to apply for this role
- Upload your resume — keep it up-to-date and in English. Mercor will auto-fill your profile from it.
- Complete the AI interview — a 15-minute conversation about your experience. Be ready to discuss specific projects and challenges you've solved.
- Submit your application — only about 20% of applicants finish all the steps, so completing yours puts you well ahead.