Anthropic is reportedly forming a dedicated team to develop custom AI chips that could eventually power future versions of its Claude models. The move comes as AI companies seek greater control over computing resources amid soaring demand for advanced processors and ongoing supply constraints dominated by firms such as Nvidia.
According to a Business Insider report, Anthropic has begun recruiting engineers with semiconductor design expertise to build an in-house silicon team. The group will focus on designing AI chips tailored specifically for Claude, with the goal of improving performance while reducing the cost of running increasingly sophisticated AI models.
The latest development builds on earlier reports that suggested Anthropic was exploring custom chip development. At the time, the initiative was still in its early stages, with no dedicated engineering team or final chip architecture in place. Reuters had also reported that Anthropic was considering partnering with Samsung to manufacture the chips, although neither company has confirmed such plans.
For now, Claude continues to run on a combination of hardware supplied by Amazon, Google and Nvidia. According to the report, Anthropic does not intend for its own chips to replace these existing partners. Instead, the company plans to continue relying on infrastructure from AWS, Google, Nvidia and AMD alongside any processors it develops internally.
Why is Anthropic developing its own chips?
The rapid advancement of AI models has significantly increased the demand for computing power, driving up costs and intensifying competition for high-performance chips.
With the supply of advanced AI processors remaining limited, many AI companies are increasingly exploring custom hardware designed specifically for their software rather than depending entirely on commercially available chips.
This combination of growing computational demands and persistent chip shortages is reportedly a key reason behind Anthropic's move into custom silicon.
Anthropic has also been expanding its AI infrastructure through other initiatives. Earlier this year, the company entered into a long-term collaboration with Google and Broadcom to work on Tensor Processing Units (TPUs), strengthening its overall computing capabilities.
Developing proprietary chips could eventually provide Anthropic with greater control over performance, operating costs and hardware availability while reducing its long-term reliance on external suppliers.
However, designing AI processors is a complex and time-consuming process. It requires specialist semiconductor engineers, manufacturing partnerships and years of testing before chips can be produced at scale.
For now, Anthropic's reported priority is building the engineering team that will lay the foundation for custom AI chips capable of supporting future generations of Claude.
