LLM Adversarial Tester & Safety Architect

Handshake AI; United States; Remote; USD 140,000 - 200,000; Full time

Job summary

Handshake AI is seeking a red-team security researcher to stress-test large language models by crafting creative adversarial prompts and multi-turn scenarios.

You will probe guardrails, study jailbreak techniques, and evaluate model responses against harm taxonomies, helping advance safer frontier AI systems. This role requires comfort with exposure to harmful content and clear written communication with research partners.

Qualifications

• Hands-on experience with multiple LLMs for testing and evaluation.
• Intuition for crafting adversarial prompts; jailbreak familiarity is a plus.
• Creative problem-solving and clear written communication skills.

Responsibilities

• Craft creative prompts and multi-turn scenarios to stress-test AI guardrails.
• Discover jailbreak, evasion and prompt-injection techniques.
• Evaluate and score model responses against structured harm taxonomies and severity rubrics.

Skills

Adversarial prompts

LLM familiarity

Safety evaluation

Written communication

Tools

ChatGPT

Claude

Gemini

Job description

Handshake AI is seeking a red-team security researcher to stress-test large language models by crafting creative adversarial prompts and multi-turn scenarios.

You will probe guardrails, study jailbreak techniques, and evaluate model responses against harm taxonomies, helping advance safer frontier AI systems. This role requires comfort with exposure to harmful content and clear written communication with research partners.