US Public Sector Compliance, Security GRC

San Francisco, CA | New York City, NY · New York City, NY · San Francisco · Seattle, WA · USD 255,000-270,000 / year · Hybrid · Posted Oct 5, 2026

At a glance

Workplace: Hybrid
Location: San Francisco, CA | New York City, NY · New York City, NY · San Francisco · Seattle, WA
Salary: USD 255,000-270,000 / year · Base salary
Visa sponsorship: Visa sponsorship available

Anthropic’s mission is to create reliable, interpretable, and steerable AI systems. We want AI to be safe and beneficial for our users and for society as a whole. Our team is a quickly growing group of committed researchers, engineers, policy experts, and business leaders working together to build beneficial AI systems.

You'll help drive the compliance programs that keep Anthropic's US government authorizations current: federal civilian, Department of War, and state and local.
This role sits in Compliance & Audit Programs and works one regime family, US public sector, from requirements through authorization and ongoing compliance.
You'll own the recurring cycles those authorizations run on and keep every frontier model's authorizations current.

Run the recurring compliance cycles for our US government authorizations (continuous monitoring, POA&M, annual assessments, SSP updates, significant changes and incident notifications), starting with C4G on Palantir FedStart and the NIST SP 800-171 assessment with Schellman
Co-own FedRAMP 20x work for Claude Enterprise, and help build new authorizations (first-party FedRAMP High, DoW impact levels, StateRAMP and TX-RAMP) with requirements and evidence work
Support model authorizations in GovCloud and Vertex for every model launch, as the GRC member of the Inference and model delivery pod
Translate government obligations into partner team requirements, for example vulnerability SLAs in our internal vulnerability reporting platform, and review the evidence
Answer public sector customer and deal questions (what a boundary allows, CUI, IRS Publication 1075, CJIS, ITAR), and handle questionnaires and RFIs until they move to Customer Trust
Map US government requirements onto the Common Control Framework with the Controls Assurance Lead and keep one source of truth for status
Build with Claude: automate mapping, evidence collection and reporting, and verify machine-drafted language before it becomes the record

Several years in security compliance or IT audit with hands-on US government compliance for a cloud service (FedRAMP, DoD impact levels, CMMC or NIST SP 800-171, or StateRAMP), including the ongoing compliance cycle after authorization
Working command of the NIST SP 800-53 Moderate baseline and the mechanics of authorization: boundary definition, control implementation statements, assessment, continuous monitoring, POA&M and significant change
Experience writing requirements for engineering teams from a control baseline and reviewing the evidence that came back
Working knowledge of how a GovCloud or Vertex style government region differs from commercial, and what changes when a model, feature or region is added to an authorized boundary
Enough technical fluency to read a runbook, configuration or pipeline definition and judge whether it enforces the written control
Clear writing, because implementation statements and status reports are what assessors, engineers and leadership work from
Ability to get partner teams to prioritize and close compliance work without authority over them

Have taken a service through FedRAMP High or DoD IL4 or IL5, or supported a service on classified networks
Have worked on a FedRAMP 20x pilot or built machine-readable evidence or reporting for an assessor
Have applied LLMs to compliance work such as control mapping, evidence testing or continuous evidence collection
Hold or are eligible for a US security clearance
Experience with state and local requirements such as StateRAMP, TX-RAMP, IRS Publication 1075 or CJIS

Location-based hybrid policy: Currently, we expect all staff to be in one of our offices at least 25% of the time.