GenAI CBRNE Cyber Expert
Summary
Red team generative AI models focusing on CBRNE and cyber threats by designing adversarial prompts, jailbreaks, and attack scenarios, then documenting failure modes and mitigation recommendations.
- Create red team test suites and evaluation rubrics
- Design adversarial prompts and jailbreaks
- Develop cyber CBRNE attack scenarios
- Evaluate model assistance to threat actors
- Execute adversarial red teaming of generative AI models
- Identify and document failure modes
- Maintain knowledge of GenAI attack techniques and cyber threat intelligence
- Produce technical documentation and mitigation recommendations
- Run multi turn model evaluations