Remote AI Red Teamer - LLM Generalist (Stress-Test)

Apply; Northern (KY); Hybrid; USD 120,000 - 180,000; Part time

Job summary

Handshake AI seeks an AI Red Teamer to stress-test large language models by crafting adversarial prompts that reveal vulnerabilities in safety guardrails, bias, and jailbreaking weaknesses. This role covers multiple risk areas including content safety, cybersecurity, and high-risk domains, across text, image, and agentic capabilities.

You will work with engineers and researchers to document experiments, score outputs against harm taxonomies, and help strengthen defenses while operating with

Qualifications

• Stress-test AI models by crafting creative adversarial prompts.
• Evaluate model outputs against harm taxonomies and severity rubrics.
• Document experiments with methods and findings.
• Collaborate with engineers and researchers to strengthen defenses.

Responsibilities

• Craft prompts and multi-turn scenarios to stress-test AI guardrails across diverse risk categories.
• Discover jailbreak, evasion, and prompt-injection techniques.
• Explore edge cases to provoke disallowed or harmful outputs.
• Evaluate and score model responses using structured harm taxonomies.

Skills

LLM experience

Adversarial prompts

Creative problem-solving

Written communication

Ethical judgment

Self-directed

Curiosity

Python basics

LLM APIs

Data annotation

Trust and safety

Subject matter expertise

Tools

Python

Job description

Handshake AI seeks an AI Red Teamer to stress-test large language models by crafting adversarial prompts that reveal vulnerabilities in safety guardrails, bias, and jailbreaking weaknesses. This role covers multiple risk areas including content safety, cybersecurity, and high-risk domains, across text, image, and agentic capabilities.

You will work with engineers and researchers to document experiments, score outputs against harm taxonomies, and help strengthen defenses while operating with