AI Safety Experts – English, Malay

This is a fully remote position, open to applicants in United States.

📋 Description

• Conduct red team assessments on conversational AI models and agents by utilizing jailbreaks, prompt injections, misuse scenarios, bias exploitation, and multi-turn manipulation techniques.

• Create human data through the annotation of failures, classification of vulnerabilities, and identification of systemic risks.

• Implement taxonomies, benchmarks, and playbooks to ensure consistent testing practices.

• Generate reproducible reports, datasets, and attack scenarios for client use.

• Investigate AI systems to reveal vulnerabilities that automated tests may overlook.

• Broaden evaluation coverage and minimize unexpected issues during production.

• Assist Mercor clients in enhancing the safety and resilience of their AI systems.

⛳️ Requirements

• Previous experience in red teaming within AI adversarial contexts, cybersecurity, or socio-technical probing.

• Capability to challenge systems to their limits through adversarial testing.

• Familiarity with frameworks or benchmarks instead of random hacking approaches.

• Proficiency in articulating risks clearly to both technical and non-technical audiences.

• Flexibility to adapt across various projects and clientele.

• Native fluency in both English and Malay.

• Experience in adversarial machine learning, including jailbreak datasets, prompt injection, RLHF/DPO attacks, or model extraction is a plus.

• Background in cybersecurity specialties such as penetration testing, exploit development, or reverse engineering is a plus.

• Expertise in socio-technical risks including harassment/disinformation probing, abuse analysis, or testing of conversational AI is a plus.

• Creative probing experience in fields such as psychology, acting, or writing is a plus.

• Candidates must not be on an H1-B or STEM OPT visa.

🏝️ Benefits

• Fully remote position.

• Flexible scheduling options.

• Participation in higher-sensitivity projects is optional.

• Access to clear guidelines and wellness resources for projects involving sensitive content.

• Weekly payments made through Stripe or Wise.

• Opportunity to gain experience in human data-driven AI red teaming.

• Play a direct role in enhancing the robustness, safety, and trustworthiness of AI systems.

• Competitive compensation.

• Collaboration opportunities with leading researchers.