AI Red Teamer, CBRNE

Evaluate AI models for CBRNE threat handling.

• •Evaluate AI models for handling queries related to CBRNE threats, ensuring they do not provide dangerous information.
• •Key Responsibilities Design adversarial prompts to test model defenses Evaluate model outputs for technical accuracy Probe dual-use knowledge boundaries Test multi-step attack chains Document findings with technical reasoning Requirements Graduate-level education in a CBRNE field Ability to evaluate model outputs Understanding of dual-use research concerns Experience with multiple LLMs Strong ethical judgment