About this role
Role Overview
Help strengthen AI systems by probing conversational models with adversarial inputs, identifying vulnerabilities, and creating actionable red-team data. This text-based remote role includes reviewing AI outputs involving sensitive subjects such as bias, misinformation, and harmful behavior. Participation in higher-sensitivity projects is optional; topics are disclosed in advance, with clear guidelines and wellness resources available.
Key Responsibilities
- Test conversational AI models and agents for jailbreaks, prompt injection, misuse, bias exploitation, and multi-turn manipulation.
- Annotate model failures, classify vulnerabilities, and flag systemic risks.
- Use taxonomies, benchmarks, and playbooks to conduct consistent testing.
- Create reproducible reports, datasets, and attack cases that support AI safety improvements.
Qualifications
- Native fluency in both English and Bengali is required.
- Prior experience in AI red teaming, cybersecurity, or socio-technical probing.
- Ability to test systems methodically, communicate risks clearly to technical and non-technical audiences, and adapt across projects.
Preferred Expertise
- Adversarial machine learning, including jailbreak datasets, prompt injection, RLHF/DPO attacks, or model extraction.
- Cybersecurity, including penetration testing, exploit development, or reverse engineering.
- Socio-technical risk analysis, including harassment or disinformation probing, abuse analysis, or conversational AI testing.
- Creative adversarial thinking informed by psychology, acting, or writing.
Work Terms
- Remote, hourly engagement.
Compensation
- $20 to $22 per hour.
Eligibility
- Applicants must be able to work remotely and meet the English and Bengali native-fluency requirement.