Remote AI Adversarial Red Teamer: Jailbreak & Prompt Injection

AI Trainer Jobs; United States; Remote; USD 96,000 - 138,000; Part time

Job summary

AI Trainer Jobs is seeking a remote contractor to perform adversarial red-teaming on AI systems. You will craft jailbreak prompts, document failures with exact reproduction steps, and assess defenses across single and multi-turn conversations.

Responsibilities include triaging new attack vectors, assigning severity, and contributing to patch tickets with clear, policy-grounded rationales. Remote work eligible from the US; hourly compensation is discussed after interview.

Qualifications

• Proven experience red-teaming AI systems or adversarial ML work.
• Ability to document failures with rigorous steps for reproduction.
• Familiarity with jailbreak, prompt-injection and policy taxonomy.
• Ability to work asynchronously for 10+ hours per week.
• Strong written communication to convert findings into patch tickets.

Responsibilities

• Design adversarial prompts that probe known weakness classes.
• Document every successful attack with reproduction steps and policy violations.
• Score model defenses across single-turn and multi-turn conversations.
• Triage emerging attack vectors and route them to the safety team.

Skills

Adversarial prompting

Red-team analysis

Policy taxonomy

Failure documentation

Jailbreak exploitation

Security review

Adversarial testing

Threat modeling

Jailbreak

Exploit

Job description

AI Trainer Jobs is seeking a remote contractor to perform adversarial red-teaming on AI systems. You will craft jailbreak prompts, document failures with exact reproduction steps, and assess defenses across single and multi-turn conversations.

Responsibilities include triaging new attack vectors, assigning severity, and contributing to patch tickets with clear, policy-grounded rationales. Remote work eligible from the US; hourly compensation is discussed after interview.