AI Adversarial Security Engineer — Jailbreak/Prompt-Injection

OpenTrain AI, Inc.; United States; Remote; USD 69,000 - 124,000; Part time

Job summary

OpenTrain AI, Inc. is seeking an AI Jailbreak and Prompt-Injection Security Expert for a part-time contractor role. You will test how AI systems respond to adversarial requests and unsafe instructions, turning security findings into practical recommendations to improve model safety.

You will design methods for testing jailbreaks, create cross-domain strategies, and build regression tests for jailbreak susceptibility, while documenting findings for both technical and non-technical audiences.

Qualifications

• At least two years in adversarial ML or AI safety evaluation.
• Professional fluency in English.
• Experience testing jailbreaks and prompt injection.

Responsibilities

• Design methods for testing jailbreaks and prompt injection.
• Create cross-domain strategies to reveal multi-turn bypass patterns.
• Build regression tests for jailbreak susceptibility.
• Develop evaluation frameworks for adversarial threats.
• Document methods and findings in reports for technical and non-technical audiences.

Skills

Adversarial ML

Red teaming

Security evaluation

Communication skills

Education

Advanced degree in CS/ML

Tools

LLM architectures

Prompt engineering

Security assessments

Job description

OpenTrain AI, Inc. is seeking an AI Jailbreak and Prompt-Injection Security Expert for a part-time contractor role. You will test how AI systems respond to adversarial requests and unsafe instructions, turning security findings into practical recommendations to improve model safety.

You will design methods for testing jailbreaks, create cross-domain strategies, and build regression tests for jailbreak susceptibility, while documenting findings for both technical and non-technical audiences.