AI safety & policy
Information security
Full-time
Routes to impact
Direct high impact on an important cause
Description
Build systems and experiments to evaluate, secure, and mitigate risks from frontier AI models.
- Create evaluation pipelines and testing frameworks for advanced models
- Build tools to analyze large transcript sets and summarize findings
- Design controlled environments and challenges for security research
- Develop agent frameworks that improve complex task completion
This text was generated by AI. If you notice any inconsistencies, please let us know using this form.



