Program Manager, Responsible Scaling Policy

About the role

We are seeking a Program Manager to own the operations behind Anthropic's Responsible Scaling Policy (RSP) - the framework that governs when our models are safe to train, deploy, and keep deployed. The RSP drives a continuous cycle of capability evaluations, Risk Reports, ASL safeguard readiness checks, Board approvals, and public commitments. You'll build and run the program infrastructure that keeps that cycle on track against an accelerating model launch cadence: source-of-truth tracking, report pipelines, and cross-functional coordination across research, Safeguards, Security, and Legal. 

This is one of the most consequential program management roles at Anthropic, sitting at the point where our safety commitments meet our models. The ideal candidate combines rigorous compliance-program instincts with the cross-functional fluency to work across research and exec stakeholders, and genuine conviction in Anthropic's safety mission.

Responsibilities

A non-exhaustive view of the work this role entails is below, including:

• Maintain the source of truth view for every model and the RSP obligations it triggers so anyone can see at a glance where things stand
• Project-manage Risk Reports and Capability Reports from end-to-end: shepherding through evals, reviews, and publication to ensure reports land when they're needed
• Pull together Safeguards, Security, and any other required reviews ahead of each launch to confirm the applicable RSP standard is met, surfacing gaps early enough to act on them
• Own the RSP commitments tracker (Frontier Safety Roadmap milestones, evaluation cadence, public update obligations) and make it the system of record the program runs on
• Run RSP version updates through internal review, Board approval, and public posting
• Represent the RSP program in launch readiness forums and system card reviews, making sure policy owners are looped in when a decision is needed

You may be a good fit if you have

• 8+ years running high-stakes, deadline-driven compliance or governance programs where missing a step had real external consequences - regulatory operations, trust & safety launch reviews, security/privacy compliance, or similar
• Direct experience as the connective tissue across technical, security, legal, and exec stakeholders who all have different vocabularies and incentives
• The ability to read a technical evaluation report, ask the right clarifying questions, and turn it into a crisp decision memo
• A demonstrated record of building tracking systems driving cross-functional initiatives to resolution
• Calm, sound judgment when launch timelines and safety obligations collide - you can hold a line on process while staying pragmatic
• Sharp written communication; your status updates and approval packets need minimal rework before going to senior leadership or externally
• Genuine investment in Anthropic's safety mission and a desire to be close to the mechanism that enforces it
• Existing context on the RSP, model launch process, or one of the contributing teams (Safeguards, Security, capability evals, Alignment Stress Testing, Policy) is a plus
• Experience preparing materials for a Board, regulator, or external auditor is a plus