AI Safety Expert – English, Punjabi

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

📋 Description

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

• Create human data through annotating failures, identifying vulnerabilities, and highlighting systemic risks.

• Implement taxonomies, benchmarks, and playbooks to maintain testing consistency.

• Generate replicable reports, datasets, and attack scenarios that clients can act upon.

• Evaluate AI outputs related to sensitive subjects like bias, misinformation, or harmful behavior.

• Discover vulnerabilities overlooked by automated testing.

• Broaden evaluation scope and minimize unexpected production issues.

• Enhance customer AI systems to improve their safety, robustness, and reliability.

⛳️ Requirements

• Proficient fluency in both English and Punjabi is essential.

• Strong discernment regarding language and content, particularly in evaluating the accuracy, completeness, and appropriateness of AI responses.

• Capability to detect subtle errors, inconsistencies, and gaps.

• Consistently adhere to guidelines and quality standards.

• Ability to articulate reasoning clearly to both technical and non-technical audiences.

• Flexibility to adapt across various projects, tasks, and clients.

• Independent contractor status required.

• Candidates must not hold H1-B or STEM OPT status, as Mercor cannot accommodate these applicants.

• Preferred: experience in adversarial machine learning, including familiarity with jailbreak datasets, prompt injection, RLHF/DPO attacks, or model extraction.

• Preferred: background in cybersecurity, such as penetration testing, exploit development, or reverse engineering.

• Preferred: experience with socio-technical risks, including harassment/disinformation probing, abuse analysis, or testing conversational AI.

• Preferred: experience in psychology, acting, or writing to foster unconventional adversarial thinking.

🏝️ Benefits

• Fully remote position.

• Flexible working hours; tasks can be completed at your convenience.

• Weekly payments through Stripe or Wise based on services provided.

• Participation in higher-sensitivity projects is optional.

• Clear content guidelines and wellness resources available.

• Reasonable accommodations offered upon request.

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

• Chance to engage in cutting-edge AI safety projects.

• Opportunity to collaborate with top researchers in the field.

• Referral bonuses of up to $90 for each successful referral, subject to limitations.