AI Safety Expert (English and Kannada) – United States Remote Contract
Mercor seeks an AI red-team specialist fluent in English and Kannada to test frontier models for vulnerabilities. The work involves creating reproducible attack cases, evaluation datasets, and risk reports for human-data AI training projects.
Description
- Test conversational AI models and agents with jailbreaks, prompt injections, misuse scenarios, bias exploitation, and multi-turn manipulation
- Create human-data annotations by documenting failures, categorizing vulnerabilities, and identifying systemic risks
- Use established taxonomies, benchmarks, and playbooks to maintain consistent evaluations
- Prepare reproducible customer-facing reports, datasets, and attack cases
- Assess AI responses on sensitive subjects, including bias, misinformation, and harmful behavior
- Investigate AI systems for weaknesses that automated testing may overlook
- Broaden evaluation coverage to help limit unexpected production issues
- Support the training and improvement of frontier models for Mercor’s AI lab and enterprise customers
Requirements
- Native-level fluency in both English and Kannada
- Sound judgment when evaluating whether AI responses are accurate, complete, and appropriate
- Clear communication skills for explaining reasoning to technical and non-technical audiences
- Meticulous attention to subtle errors, inconsistencies, and omissions
- Consistency in applying taxonomies, benchmarks, playbooks, guidelines, and quality standards
- Flexibility to work across different projects, task formats, and customers
- Ability to work as an independent contractor
- Must not require H-1B or STEM OPT sponsorship; these sponsorship pathways are not supported
- Relevant experience in adversarial machine learning, cybersecurity, sociotechnical risk, conversational AI testing, psychology, acting, or unconventional adversarial writing is advantageous
Benefits
- Work fully remotely with the flexibility to set your own schedule
- Receive weekly payments through Stripe or Wise for services provided
- Choose whether to participate in projects involving higher-sensitivity content
- Access clear guidance and wellness resources for sensitive-content assignments
- Receive competitive compensation
- Develop practical experience in human-data-driven AI red teaming
- Collaborate with leading researchers
- Request reasonable accommodations when needed
- Earn referral payments of up to $90 for each successful referral