Staff + Software Security Engineer, Labs

Details

Anthropic’s Security Labs team runs high-risk, high-expected-value security projects for frontier AI systems. Projects typically run for weeks and are either handed off to the Anthropic team that will own them in production or wound down with a documented writeup of the results.

The team works on problems including extreme isolation for research workflows, cryptographic guarantees for model execution, hypervisor-isolated workloads, formally verified compilation, regenerating clusters, and using AI systems to support security work. Engineers help select projects and are expected to work independently in uncertain, rapidly changing environments.

Responsibilities

• Own Security Labs projects end to end, including scoping the project, building prototypes, testing them against real workloads, and handing them off or documenting an exit.
• Stand up novel security infrastructure, including isolated clusters, attestation chains, hypervisor and runtime systems, and verification tooling.
• Identify receiving teams early and build solutions they can adopt and own.
• Work with Pretraining, RL, Inference, and Compute teams to test proposed systems against real workflows.
• Turn experimental results into concise technical writeups and costed contingency plans.
• Help select future security projects and influence broader industry security practices.

Requirements

• Strong interest in the security challenges affecting the future of AI.
• Deep expertise in at least one specialized area, such as firmware or hardware security, applied cryptography, OS, kernel or hypervisor internals, formal methods and verification, reverse engineering and exploit development, or high-assurance and cross-domain systems.
• A record of building and shipping projects independently, such as founding a company or research group, maintaining a relied-upon open-source project, or producing influential research.
• Experience choosing and pursuing problems independently, working across domains and stack layers, and running prototypes or experiments, including projects that did not succeed.
• Strong written communication skills and the ability to document technical results clearly.
• A defensive security focus, although offensive security, red teaming, and vulnerability research experience is valuable.
• Strong programming skills in Python plus at least one of Rust, Go, or C/C++, with the ability to build real infrastructure.

Preferred Qualifications

• Experience in airgapped or high-side environments, including classified networks, cross-domain solutions, ICS/SCADA, or financial trading infrastructure.
• Background in offensive security, red teaming, or vulnerability research.
• Familiarity with ML infrastructure, including training pipelines, distributed schedulers, inference serving, and accelerator hardware.
• Experience in rapidly iterating environments such as startups, applied research groups, independent consulting, or small security teams.

Education and Logistics

A bachelor’s degree or equivalent combination of education, training, and experience is required. The field of study must be relevant to the role through coursework, training, or professional experience. Required experience correlates with the internal job level.

Anthropic expects staff to work from one of its offices at least 25% of the time, though some roles may require more office time. Anthropic explicitly sponsors visas and will make reasonable efforts to assist with visa applications, supported by an immigration lawyer.

Anthropic offers competitive compensation and benefits, optional equity donation matching, generous vacation and parental leave, flexible working hours, and an office environment for collaboration.