Engineering Manager, Hardware Platform Security

Leads Anthropic's Platform Security Engineering group, part of Data Center Security Engineering, and reports to that function's manager. As Anthropic moves workloads beyond the big cloud providers into its own facilities, colocation and partner sites, the group takes on the security that clouds used to handle for it, so training and inference can run anywhere with a provable security posture, including the technical controls behind its AI Safety Level commitments.

The manager hires and develops engineers whose work spans firmware research, kernel and operating system hardening, platform trust integration and server management firmware, and coordinates the contractor and vendor engineers who work with them. The role owns the roadmap for hardware roots of trust and attestation, secure and measured boot for CPUs, BMCs, accelerators and peripherals, and the host-side process that uses attestation results to decide what reaches production. It also owns the fleet's firmware supply chain posture, runs risk assessments of new platforms and providers with go or no-go recommendations, manages security work with silicon vendors, OEMs and compute providers, and engages standards bodies such as TCG, OCP and IETF. It is hands-on in the sense of reviewing designs and settling hard trade-offs rather than writing most of the code.

Anthropic wants ten or more years in systems security, five of them on firmware, hardware or OS security, and five or more managing security or systems engineers, including senior and staff engineers. Fluency in secure boot, TPMs, SPDM and DICE attestation, UEFI, BMC firmware and Linux hardening is expected. Based in San Francisco or Seattle.