AI Safety Expert - Adversarial ML

About the role

Mercor connects elite creative and technical talent with leading AI research labs. Headquartered in San Francisco, our investors include Benchmark, General Catalyst, Peter Thiel, Adam D'Angelo, Larry Summers, and Jack Dorsey. This is a contract position, remote, with compensation of $16–$22/hour.

Responsibilities

• Red team conversational AI models and agents.
• Perform jailbreaks, prompt injections, misuse cases, and bias exploitation.
• Generate high-quality human data.
• Annotate failures, classify vulnerabilities, and flag systemic risks.
• Apply structure by following taxonomies, benchmarks, and playbooks to ensure consistent testing.
• Document reproducibly to produce reports, datasets, and attack cases that customers can act on.
• Work independently and asynchronously to meet deadlines while improving AI model performance.

Qualifications

Must‑have:

• Native fluency in English and Bengali.
• Strong judgment about language and content accuracy.
• Rigorous attention to subtle errors and inconsistencies.
• Ability to work to guidelines and quality standards consistently.
• Clear communication with technical and non‑technical audiences.
• Adaptability across projects, task types, and customers.

Preferred:

• Experience in Adversarial ML: jailbreak datasets, prompt injection, RLHF/DPO attacks, model extraction.
• Cybersecurity skills: penetration testing, exploit development, reverse engineering.
• Socio‑technical risk expertise: harassment/disinfo probing, abuse analysis, conversational AI testing.
• Creative probing skills: psychology, acting, writing for unconventional adversarial thinking.