Program Manager, Safeguards Policy, Enforcement, and Threat Intelligence

Anthropic builds AI systems that are safe, reliable, and interpretable. The company's mission centers on creating beneficial AI that serves users and society well. With a fast-growing team of researchers, engineers, policy experts, and business leaders, Anthropic is at the forefront of developing AI systems that balance capability with responsible deployment.

The Safeguards organization at Anthropic is responsible for defining and enforcing the boundaries of how Claude can be used. This involves three interconnected functions: Policy Design establishes what is and isn't permitted across different harm categories, Enforcement detects and responds to violations in real time, and Threat Intelligence investigates sophisticated misuse patterns and emerging risks. These teams move quickly and their decisions ripple across the company, affecting model training, detection systems, product launches, legal considerations, and external messaging.

This Program Manager role will serve as the operational backbone for the Safeguards organization. You'll maintain visibility across all three teams, establish and run the processes that connect their work to each other and to the broader company, and ensure alignment before major decisions take effect. You'll also manage the planning cadence, track priorities and status, monitor decision logs, and coordinate readiness for policy and enforcement work tied to model releases.

What you'll do

• Lead cross-functional program management across Policy Design, Enforcement, and Threat Intelligence, keeping all workstreams visible to leadership and on schedule
• Establish and run the operating rhythm including quarterly planning, priority setting, status updates, and documented decision records
• Coordinate policy initiatives with research, product, engineering, legal, and communications teams to surface dependencies early
• Manage readiness processes for policy and enforcement work supporting model launches and product releases, including timelines, reviews, and approvals
• Own the external publication process for blog posts, threat reports, and other communications from conception through publication
• Identify risks, blockers, and resource gaps across the portfolio and propose solutions to leadership
• Design lightweight processes and tools that improve team velocity, and eliminate processes that become overhead

What you'll bring

• Demonstrated experience running cross-functional programs in fast-moving technical environments with a track record of shipping on time across teams you don't directly manage
• Ability to take ambiguous goals and break them into clear plans, owners, and milestones that you keep current as circumstances change
• Strong written communication skills including clear status updates, decision documentation, and executive summaries
• Experience managing review and approval processes with multiple stakeholders in contexts like product launches, policy releases, or publications
• Sound judgment about when process adds value versus when it creates overhead, with a preference for the lightest process that works
• Comfort handling sensitive topics and maintaining strict confidentiality standards

Nice to have

• Background in trust and safety, product policy, enforcement operations, threat intelligence, or AI safety
• Understanding of how AI models are developed and launched, and how policy and enforcement integrate into that cycle
• Experience coordinating external research communications or public-facing releases
• Experience using AI tools to automate tracking, reporting, and recurring program management tasks

What they offer

• Annual salary range of 285,000 to 330,000 USD
• Visa sponsorship support for qualified candidates
• Hybrid arrangement with expectation of in-office presence at least 25% of the time, based in San Francisco