AI Red Team - Adversarial AI Testing (Novice) Job

Mercor is hiring AI Red-Teamers to probe large language models and agents with adversarial prompts, jailbreaks, misuse cases, and exploits to surface vulnerabilities. You will annotate failures, classify risks, and document reproducible attack cases that strengthen AI safety. Work includes red-teaming across sensitive domains (bias, misinformation, misuse) with clear guidelines and wellness support. Projects vary across AI jailbreaks, socio-technical probing, and adversarial ML tasks. This role is fully remote and available part-time or full-time (US time zones preferred).

Apply here: https://work.mercor.com/jobs/list_AAABmYNIhHBkrcVkMSlCRYu8?referralCode=0b7b9943-2751-458b-bbda-261a70593f6e&utm_source=referral&utm_medium=share&utm_campaign=job_referral

Prior red-teaming, cybersecurity, adversarial ML, or socio-technical probing experience preferred.

Strong curiosity and adversarial mindset, with ability to push systems to their limits.

Structured, framework-driven approach to testing.

Excellent communication skills to report risks clearly.

Adaptability to work across multiple projects and customer needs.

Based in the US, UK, or Canada.

Flex Time, Other Benefits

Mercor connects elite creative and technical talent with leading AI research labs worldwide. Headquartered in San Francisco, Mercor partners with top institutions to advance cutting-edge projects in AI, science, and engineering. Our distinguished investors include Benchmark, General Catalyst, Peter Thiel, Adam D’Angelo, Larry Summers, and Jack Dorsey. Thousands of professionals across domains such as law, engineering, and research collaborate with Mercor to shape the future of AI innovation.