Anthropic Is Building an In-House Chip Team to Design Custom AI Silicon for Claude

Anthropic is assembling an in-house chip design team to develop custom AI accelerators for its Claude family of models, according to multiple reports. The move makes Anthropic the latest AI company to pursue vertical integration in silicon, following efforts by Google, Amazon, and Microsoft.

Moving Beyond Third-Party Dependence

Like most AI companies, Anthropic currently relies on NVIDIA GPUs to train and run its large language models. But as model sizes continue to grow, compute costs for training and inference have become one of the largest expenses for AI labs. Custom silicon allows companies to optimize hardware architectures for specific workloads, achieving better performance and lower costs for their particular use cases.

The decision reflects a broader industry trend: leading AI labs are increasingly moving toward vertical integration, extending their reach from model development down to the underlying hardware layer to reduce dependence on a limited number of chip suppliers.

Hiring Chip Design Engineers

According to reports from Reuters and Business Insider, Anthropic has already begun recruiting chip design engineers. While the company has not disclosed a timeline or technical specifications for its chip project, building a dedicated team signals a long-term strategic commitment rather than a short-term experiment.

Anthropic’s main rival, OpenAI, is also exploring custom chip development and has reportedly held discussions with semiconductor firms including Broadcom and TSMC. This suggests the AI chip market may see a wave of customized solutions from AI companies in the coming years.

The AI Chip Race Intensifies

The market for AI-specific chips is becoming increasingly competitive. NVIDIA currently dominates the data center AI chip market, but AMD and Intel are accelerating their efforts, while cloud providers like Google and Amazon have already deployed multiple generations of custom AI silicon. Anthropic’s entry further validates that custom chip design is becoming standard practice for top AI companies.

For Anthropic specifically, custom silicon offers more than cost savings. Controlling the hardware layer could provide greater oversight for Claude’s safety research, a strategic advantage for a company whose core mission centers on AI safety.

Source: Reuters · Business Insider