Cybersecurity SWE — AI Safety

AI Trainer Jobs; United States; Remote; USD 83,000 - 124,000; Full time

Job summary

AuraOne is seeking Cybersecurity SWE — AI Safety for a remote red-team track to stress-test AI systems against adversarial prompts. Reviewers craft attack scenarios, document failure modes, and pair each jailbreak with the violated rubric clause to patch gaps.

Adversarial evaluation helps harden models before shipping. Candidates think like attackers, write rigorous failures for reproducibility, fixing, and regression testing; push on refusal boundaries and dual-use risk before release.

Qualifications

• Experience red-teaming AI systems or adversarial ML.
• Strong written communication; reports become patch tickets.
• Comfort documenting policy-grey areas with clear reasoning.

Responsibilities

• Design adversarial prompts to test known weakness classes.
• Document successful attacks with reproduction steps and violated policy clause.
• Score model defenses across single-turn and multi-turn conversations.
• Triage emerging attack vectors and route to safety team with severity ratings.
• Maintain a library of attack patterns and propose new red-team rubrics.

Skills

Red-teaming

Adversarial ML

Security research

Technical writing

Policy understanding

Multilingual testing

Job description

Cybersecurity SWE — AI Safety is a remote red-team track for stress-testing AI systems against adversarial prompts. Reviewers craft attack scenarios, document the failure mode, and pair each successful jailbreak with the rubric clause it violated so the safety team can patch the gap.

Category: AI Safety & Red Teaming · Pay: $60–$90 / hr · Location: Remote — US-eligible · Contractor

Cybersecurity SWE — AI Safety is a remote red-team track for stress-testing AI systems against adversarial prompts.

Role details

About the role

Cybersecurity SWE — AI Safety is a remote red-team track for stress-testing AI systems against adversarial prompts. Reviewers craft attack scenarios, document the failure mode, and pair each successful jailbreak with the rubric clause it violated so the safety team can patch the gap.
Adversarial evaluation is how AuraOne hardens AI models before they ship to customers. Reviewers think like attackers and write up failures with enough rigor that the modeling team can reproduce, fix, and regress-test them.
Push on refusal boundaries and dual-use risk before a model ships.

Responsibilities

• Design adversarial prompts that probe known weakness classes (jailbreak, policy bypass, prompt injection) for Cybersecurity SWE — AI Safety assignments.
• 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.
• Maintain a personal library of attack patterns and propose new red-team rubrics.

What you should bring

• Demonstrated experience red-teaming AI systems, security research, or adversarial ML work for Cybersecurity SWE — AI Safety work.
• Strong written communication — your reports become the patch ticket.
• Comfort working in policy-grey areas with clear documentation of what was attempted and why.
• Familiarity with prompt-injection, jailbreak, and policy-bypass taxonomies.
• Reliable async availability for at least 10 hours per week.

Example tasks

• Construct a 5-turn adversarial conversation that bypasses a specific policy clause and write up the patch ticket.
• Score a model's defenses against a known jailbreak pattern across 20 variants.
• Propose a new red-team rubric category after spotting an emerging attack vector.
• Reproduce a failure another reviewer reported and confirm the severity tag.

Useful experience

• Background in offensive security, AppSec, or trust & safety operations.
• Experience publishing or reproducing public adversarial-ML research.
• Multilingual fluency for cross-language attack testing.

Compensation and schedule

$60–$90 / hr
Expected arrangement: contractor , with program-defined task volume and review pacing. Placement depends on current program demand and reviewer confirmation.

Skills used in matching

• Adversarial prompting
• Red-team analysis
• Policy taxonomy
• Failure documentation
• Cybersecurity SWE AI Safety evaluation