Remote AI Safety Red-Team Expert (Turkish, PhD)

AuraOne; United States; On-site; USD 41,000 - 47,000; Full time

Job summary

AuraOne is seeking a Bilingual Turkish STEM Expert (PhD) for a remote AI Safety red-team track. You will design adversarial prompts, document failures with rigor, and pair jailbreaks with rubric violations to help patch safety gaps.

The role emphasizes asynchronous collaboration and high-quality incident reporting. Responsibilities include scoring defenses in both single-turn and multi-turn interactions, triaging new attack vectors, and maintaining a personal library of attack patterns.

Qualifications

• Demonstrated experience red-teaming AI systems, security research, or adversarial ML work.
• Experience working on AI safety evaluations, preferably with Turkish language scope.
• Strong reporting and documentation skills to accompany patches.

Responsibilities

• Design adversarial prompts that probe known weaknesses (jailbreak, policy bypass).
• Document every successful attack with reproduction steps and the violated policy clause.
• Score model defenses across single-turn and multi-turn conversations.
• Triadge emerging attack vectors and route them to the safety team with severity ratings.
• Maintain a library of attack patterns and propose new red-team rubrics.
• Coordinate with the broader red-team cohort to ensure consistent coverage.

Skills

Adversarial prompting

Red-team analysis

Policy taxonomy

Failure documentation

Turkish generalist evaluation

Job description

AuraOne is seeking a Bilingual Turkish STEM Expert (PhD) for a remote AI Safety red-team track. You will design adversarial prompts, document failures with rigor, and pair jailbreaks with rubric violations to help patch safety gaps.

The role emphasizes asynchronous collaboration and high-quality incident reporting. Responsibilities include scoring defenses in both single-turn and multi-turn interactions, triaging new attack vectors, and maintaining a personal library of attack patterns.