AI safety & policy
Full-time
Routes to impact
Skill-building & building career capital
Direct high impact on an important cause
Description
Design and run red teaming tests for generative AI systems by turning emerging risks, news, and regulation into practical adversarial scenarios.
- Create jailbreaks, adversarial prompts, and scaled test datasets across text, image, video, and multimodal systems
- Track frontier AI risks, current events, and regulation to develop structured testing scenarios
- Map real-world misuse risks to specific AI products and features
- Help design testing methodologies and review deliverables for quality
This text was generated by AI. If you notice any inconsistencies, please let us know using this form.
