AI Safety Red Team Expert, English and Tamil

Help identify weaknesses in conversational AI by testing model responses, uncovering risks, and creating actionable safety data in both English and Tamil. Role Overview

This remote, text-based role focuses on adversarial testing of AI models and agents. You will evaluate outputs for issues involving bias, misinformation, and harmful behavior. Participation in projects involving higher-sensitivity content is optional; topics are communicated in advance, with clear guidelines and wellness resources available.

• Test conversational AI models and agents for jailbreaks, prompt injections, misuse cases, bias exploitation, and multi-turn manipulation.
• Review AI responses for accuracy, completeness, appropriateness, and safety concerns.
• Annotate failures, classify vulnerabilities, and flag systemic risks.
• Follow established taxonomies, benchmarks, playbooks, and quality standards to conduct consistent testing.
• Create reproducible reports, datasets, and attack cases that teams can use to strengthen AI systems.

• Native fluency in English and Tamil is required.
• Strong judgment when assessing language and content, with the ability to explain findings clearly to technical and non-technical audiences.
• Careful, structured work habits and the ability to identify subtle errors, inconsistencies, and gaps.
• Adaptability across projects, task types, and customer needs.

• Adversarial machine learning, including jailbreak datasets, prompt injection, RLHF or DPO attacks, or model extraction.
• Cybersecurity, including penetration testing, exploit development, or reverse engineering.
• Socio-technical risk assessment, including harassment or misinformation testing, abuse analysis, or conversational AI testing.
• Creative adversarial thinking informed by psychology, acting, or writing.

• Remote, hourly engagement.

• $16 to $22 per hour.