Anthropic is establishing an internal chip design team to develop custom silicon for its Claude artificial intelligence models, signaling a strategic push to gain greater control over the hardware powering its systems.
The company has begun recruiting engineers with expertise in processor architecture and semiconductor design, marking a significant step toward reducing its reliance on external chip suppliers as demand for AI computing capacity continues to surge.
Custom silicon has become a critical priority across the AI industry, where the cost and availability of high-performance processors directly shape how quickly companies can train and deploy their models. By developing hardware tailored to Claude’s specific requirements, Anthropic aims to improve efficiency and potentially lower the substantial expense of running large language models at scale.
The move places Anthropic alongside a growing number of technology firms pursuing in-house chip capabilities rather than depending entirely on merchant suppliers. Building bespoke processors allows companies to optimise performance for their own workloads, an approach that has gained traction as AI systems grow larger and more computationally demanding.
Anthropic, founded by former OpenAI researchers, has rapidly emerged as one of the leading developers of frontier AI systems. Its Claude family of models competes directly with offerings from OpenAI and Google, and the company has expanded its product range to include tools capable of handling complex, multi-step tasks.
The expansion into chip design comes amid a period of aggressive growth for the company. Anthropic has been in discussions over a major funding round that would value the firm at roughly $170 billion, underscoring investor appetite for companies at the forefront of generative AI.
Assembling a competitive chip team is a formidable undertaking. The semiconductor sector faces an acute shortage of skilled engineers, a constraint that has intensified competition for talent among established chipmakers and newer entrants alike. Anthropic will need to attract specialists in an increasingly crowded market to realise its hardware ambitions.
Developing custom silicon typically requires years of investment and collaboration with foundries and design partners before production chips reach data centres. The initiative reflects a longer-term bet that owning more of the technology stack will prove essential to sustaining performance gains and managing costs.
This development comes as the broader AI industry confronts mounting pressure over the scarcity and price of advanced processors, with several major players seeking to secure supply through direct hardware development. Whether Anthropic’s in-house effort ultimately reduces its dependence on external suppliers will depend on how quickly it can build the team and bring designs to fruition.