AI Safety Expert for English and Bengali

Lead adversarial testing of conversational AI in English and Bengali, producing reproducible attack cases, datasets, and reports that surface bias, misinformation, and harmful behavior so customers can make their models safer. This is a text-only, remote role focused on human-driven red teaming and vulnerability discovery.

• Red team conversational AI models and agents, including jailbreaks, prompt injections, misuse cases, bias exploitation, and multi-turn manipulation.
• Generate high-quality human data by annotating failures, classifying vulnerabilities, and flagging systemic risks.
• Follow and apply taxonomies, benchmarks, and playbooks to keep testing consistent and comparable across projects.
• Document attacks and findings reproducibly, producing reports, datasets, and concrete attack cases that clients can act on.
• Optionally participate in higher-sensitivity projects, following provided guidelines and using available wellness resources.
• Work only with text-based content, with topics communicated in advance of exposure.

• Native fluency in English and Bengali is required.
• Prior red teaming experience, such as AI adversarial testing, cybersecurity, or socio-technical probing.
• Curious and adversarial mindset, with an instinct to push systems to breaking points.
• Structured approach to testing, using frameworks or benchmarks rather than ad hoc methods.
• Strong communication skills, able to explain risks to both technical and non-technical stakeholders.
• Adaptable to moving across projects and client contexts.

• Adversarial ML experience, including jailbreak datasets, prompt injection, RLHF/DPO attacks, or model extraction.
• Cybersecurity skills such as penetration testing, exploit development, or reverse engineering.
• Socio-technical risk expertise, including harassment, disinformation, or conversational abuse analysis.
• Creative probing skills, for example psychology, acting, or writing that supports unconventional adversarial thinking.

• Identify vulnerabilities that automated tests miss.
• Deliver reproducible artifacts that materially strengthen client AI systems.
• Expand evaluation coverage so more scenarios are tested and fewer surprises occur in production.
• Increase client trust in the safety and robustness of their AI through adversarial verification.

• Build hands-on experience in human data-driven AI red teaming at the frontier of safety.
• Make a direct impact on the robustness, safety, and trustworthiness of deployed AI systems.

• Remote role, work from your location.
• Employment type: hourly.
• All work is text-based.
• Participation in higher-sensitivity content is optional and supported by clear guidelines and wellness resources; topics will be communicated before exposure.

• Pay rate: $20 to $22 per hour.

• Native fluency in both English and Bengali is mandatory.
• Remote work capability is required.