Anthropic Fellows Program, AI Security
Core
Empirical AI research project focused on AI security, vulnerability research, and reducing catastrophic risks from advanced AI systems.
Role type
Research Fellow (AI Security)
Builds
Public research outputs (e.g., papers) and security analysis
Domain
AI Safety & Security
Deliverable
research
Required skills
Python programming, empirical ML research, deep learning frameworks, experiment management, pentesting, vulnerability research, offensive security
Preferred skills
Open-source contributions in LLM/security, CVE reporting, bug bounty awards, clarity in ambiguous technical problems
Technologies
Python, deep learning frameworks
Responsibilities
Conduct empirical research on AI security vulnerabilities, perform pentesting and vulnerability research, produce public research outputs, collaborate with mentors on security projects
Seniority
Early-career researcher (no prior experience required)