AI Safety Red Team Expert, English & Urdu

Role Overview

Help strengthen conversational AI systems by testing them from an adversarial perspective. You will probe models and agents for vulnerabilities, create high-quality safety data, and produce practical findings that support more robust and trustworthy AI.

Key Responsibilities

• Test conversational AI models and agents for jailbreaks, prompt injections, misuse cases, bias exploitation, and multi-turn manipulation.
• Review text-based AI outputs involving sensitive topics, including bias, misinformation, and harmful behaviors.
• Annotate failures, classify vulnerabilities, and flag systemic risks.
• Follow established taxonomies, benchmarks, and playbooks to ensure consistent testing.
• Create reproducible reports, datasets, and attack cases that teams can use to improve AI systems.

Qualifications

• Native fluency in both English and Urdu is required.
• Strong judgment about language and content, including the ability to assess whether an AI response is accurate, complete, and appropriate and explain why.
• Exceptional attention to subtle errors, inconsistencies, and gaps.
• Ability to work consistently within guidelines and quality standards.
• Clear communication skills for technical and non-technical audiences.
• Adaptability across projects, task types, and customer needs.

Preferred Experience

• 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 work, such as harassment or misinformation probing, abuse analysis, or conversational AI testing.
• Creative adversarial thinking informed by psychology, acting, or writing.

Work Terms

• Remote, hourly engagement.
• All work is text-based.
• Higher-sensitivity projects are optional and include clear guidelines and wellness resources. Topics will be communicated before any exposure to content.

Compensation

• $16 to $22 per hour.