Red Team Engineer

About the Role

Gray Swan Arena is where the world's top red-teamers test frontier AI models, learn about AI security, and win cash prizes. Leading labs use the resulting data to ensure that future LLMs are safe and aligned with human values. Gray Swan's Red Team Engineers are the hands-on practitioners behind both our popular Arena challenges and private red team engagements. You'll design realistic attack scenarios, find vulnerabilities in frontier AI systems, and translate your findings into clear and actionable reports.

What You'll Do

• Design new challenges and adversarial scenarios for ongoing and upcoming Arena competitions
• Work red teaming engagements with external clients, covering areas like prompt injection, tool use, and agent exploitation
• Use Slack, Linear, our Arena design editor, and AI tools to deliver red-teaming results on time and with high quality
• Build internal tools and processes for the Arena (e.g., cheating detection) and support operational tasks like break verification and prize distribution
• Work alongside Product and Engineering teams to create realistic scenarios that highlight attack surfaces in frontier systems
• Help shape the future of Gray Swan Arena with novel ideas and challenge types

Who You Are

• 1+ years of experience in AI red teaming, adversarial ML, or related security research
• Organized and comfortable working independently across multiple concurrent engagements
• Proven ability to convey technical ideas in accessible reports
• Comfortable with quick pivots and changing priorities
• Experience as a Gray Swan Arena competitor is a plus
• Background in traditional security research, CTFs, or penetration testing

Bonus Points

• Experience with AI red teaming beyond LLM prompt injection
• Degree or experience equivalent in Computer Science or related field
• Familiarity with multimodal attack surfaces
• Coding experience for automating or scaling test cases

If you don't have 100% of these, you should still seriously consider applying. We care more about what you can do than your credentials.

You'll Thrive Here If You

• Have writing and scenario design skills with a collaborative communication style for effective cross-functional teamwork
• Have strong intuition about how AI systems fail in real-world deployments
• Take ownership and deliver high-quality work under deadlines
• Want to be on the cutting edge of AI safety and security, proactively defending against novel threat vectors

Pay

$110,000 - $290,000 (depending on level) plus performance based bonus and meaningful equity package

Benefits

• 401k with up to 4% matching
• 28 days annual leave (vacation + holidays)
• Health, dental, and vision coverage
• Catered lunches (Pittsburgh office)
• Flexible work arrangements
• Visa sponsorship available for exceptional candidates