Anthropic builds in-house silicon team to design custom AI chips

Anthropic is officially entering the custom silicon race. The Claude developer is hiring engineers with direct experience shipping semiconductor designs, as first reported by Business Insider. The job listing targets experts across chip design and verification, with salaries ranging from $320,000 to $485,000 — a clear premium for talent who can take processors from research to production. Anthropic confirmed the team is part of a broader "multi-chip" strategy that will complement, not replace, its existing partnerships with AWS, Google, Nvidia, and AMD. The company hasn't disclosed a timeline for its first chip or whether it will manufacture in-house, though June reports suggested talks with Samsung Electronics about co-developing custom AI processors.

The move echoes Apple's Apple Silicon playbook: designing hardware specifically for Claude's workloads could improve performance, efficiency, and cost predictability while reducing dependence on a tight AI chip supply. Anthropic is racing to secure compute capacity as demand surges — a problem highlighted by Moonshot AI pausing signups for its Kimi K3 model due to infrastructure limits. For end users, purpose-built chips may eventually deliver faster, cheaper AI services, but immediate effects are unlikely. More broadly, this signals that AI leaders now compete on infrastructure, not just model quality, with vertical integration becoming a key strategic lever.