Remote AI Safety Adversarial Red-Teamer (Dutch)

AuraOne; United States; A distancia; USD 66.000 - 72.000; Jornada completa

Descripción de la vacante

AuraOne is seeking a Bilingual Dutch (Belgium) Generalist Expert — AI Safety contractor to design adversarial prompts and document failures with rigorous repro steps. You will help harden AI models by scoring defenses and triaging new attack vectors, ensuring patches address specific policy violations.

The role emphasizes clear, evidence-based write-ups and collaboration with the safety team to patch gaps before deployment.

Formación

• Demonstrated red-teaming experience for AI systems or adversarial ML work.
• Strong written documentation of failures and patches for reproducibility.
• Comfort working in policy-grey areas with clear documentation of what was attempted and why.

Responsabilidades

• Design adversarial prompts to probe known weakness classes (jailbreak, policy bypass, prompt injection) for Dutch Belgium Generalist Expert — AI Safety work.
• Document every successful attack with reproduction steps and the policy clause it violated.
• Score model defenses across single-turn and multi-turn conversations.
• Triage emerging attack vectors and route them to the safety team with severity ratings.

Conocimientos

Red-team testing

Adversarial ML

Policy taxonomy

Dutch (Belgium)

Technical writing

Descripción del empleo

AuraOne is seeking a Bilingual Dutch (Belgium) Generalist Expert — AI Safety contractor to design adversarial prompts and document failures with rigorous repro steps. You will help harden AI models by scoring defenses and triaging new attack vectors, ensuring patches address specific policy violations.

The role emphasizes clear, evidence-based write-ups and collaboration with the safety team to patch gaps before deployment.