Anthropic has begun assembling an internal team to design custom chips for Claude, as the AI industry accelerates efforts to build more of its critical computing infrastructure internally.
The AI developer disclosed the move through a new Silicon Engineer recruitment notice, saying it is expanding work conducted with semiconductor partners by establishing its own custom-chip group.
The team will collaborate with Anthropic’s inference, performance, infrastructure and software-kernel specialists to align future hardware more closely with Claude’s workloads.
Anthropic targets every stage of chip development
Anthropic is recruiting across front-end chip design, pre-silicon verification, physical design, testing, analog systems, foundry technology, packaging and design infrastructure, signaling that the effort will cover much of the semiconductor development cycle rather than focus on a single component.
Engineers will help set specifications, shape chip architecture and decide which technologies Anthropic should build, purchase or license.
The company also expects the team to work with outside chip designers, intellectual-property suppliers, foundries and testing partners, indicating that internal design authority will remain closely connected to external manufacturing expertise.
The position includes responsibility for bringing up and debugging the first completed silicon, although Anthropic has not disclosed a launch timetable, manufacturing partner, investment figure or technical architecture.
Bringing chip design closer to Claude’s model and software teams could allow Anthropic to tailor performance, memory use and power efficiency around its own training and inference needs. However, the initiative does not represent a retreat from the company’s existing suppliers.
A new layer in a diversified hardware strategy
Anthropic currently trains and operates Claude across Amazon Web Services’ Trainium processors, Google’s tensor processing units and Nvidia graphics chips, describing that mix as a way to match workloads with suitable hardware while improving resilience.
In April, Anthropic committed more than $100 billion over 10 years to AWS technology under an agreement securing as much as five gigawatts of capacity, spanning Trainium2 through Trainium4 and potentially later Amazon chips.
Washington and Beijing bring chipmaking closer to home
Anthropic’s move comes as companies and governments push to localize more of the semiconductor supply chain.
On July 21, the first major sections of ASML’s next-generation High-NA EUV system arrived at New York’s Albany NanoTech Complex, beginning installation of a publicly accessible research facility expected to start advanced chip work in early 2027.
China followed with its own localization push, moving a domestically engineered immersion DUV lithography system into production under Shanghai Aishengna.
The platform remains behind ASML’s established machines and still requires further validation, but even limited domestic output could give Chinese chipmakers a strategic alternative if Western restrictions on equipment, maintenance or technical support tighten.
AI reshapes national priorities and the workforce
The AI boom is reshaping government priorities, turning chip capacity, data centers and workforce skills into matters of economic security.
It is also disrupting labor markets, with the ILO estimating that one in four jobs is exposed to generative AI, although transformation remains more likely than outright replacement.
At the same time, AI is creating new roles in engineering, infrastructure and digital services, suggesting the shift will reshape employment rather than simply eliminate jobs, and could reward countries that move quickly to expand local capacity and retrain workers.
