Frontier AI Safety Red Team Expert

Obsidian is seeking experienced AI Safety Red Teamers to identify vulnerabilities in frontier AI systems through adversarial testing, prompt design, and rigorous model evaluation. You will simulate jailbreaks, assess unsafe behaviours, and document weaknesses to inform safety benchmarks and red-teaming reports, collaborating with researchers to improve alignment and robustness.

This role requires strong analytical writing and a track record in cybersecurity, safety, or life sciences, with at

We would love to welcome a new Frontier AI Safety Red Team Expert to our team in New York, NY, United States.

For the Frontier AI Safety Red Team Expert position at Mercor, we are reviewing applications now.

Step into the Frontier AI Safety Red Team Expert role at Mercor in New York, NY, United States and grow with us.

Please review the full job details above before applying.