Security Engineer, Offensive Security

Offensive security engineer on Anthropic's Security Engineering team who runs red and purple team engagements that simulate advanced threat actors against cloud infrastructure, endpoints and bare-metal deployments. The engineer also penetration tests specific high-value deployments, helps build AI-assisted security testing tools, works with other security and engineering teams on attack scenarios specific to AI, and presents findings to technical and executive audiences as risk insights that feed security roadmaps.

The posting asks for five or more years of hands-on red teaming and offensive operations, deep expertise in at least two of macOS, Linux, cloud, Kubernetes and CI/CD security, adversary simulation against well-defended environments, strong Python or Go, and close collaboration with the blue team. Remote-friendly in the US.