Anthropic is exploring the possibility of designing its own AI chips, per a report from Reuters.

Citing three unnamed sources, the move is said to be a response to the shortage of AI chips needed by the company to further power and develop its Claude AI model.

The report went on to note, however, that the discussions were still in the early stages and the generative AI company has yet to put a team together for the project or commit to a design. Anthropic could ultimately decide just to purchase additional AI chips rather than moving ahead with its own designs, the sources said.

Anthropic declined Reuters ’ request for comment.

The report comes as Anthropic announced it had signed a deal with Broadcom and Google for the supply of 3.5GW-worth of Google Tensor Processing Units (TPUs), slated to come online from 2027. In a statement, the company said the “vast majority of the new compute” would be situated in the United States and would form part of the company’s November 2025 pledge to invest $50bn in strengthening US compute infrastructure.

Also last week, AI cloud provider CoreWeave said it would be providing Anthropic with cloud compute capacity under a multi-year deal, although the value of the agreement or details about the hardware/instances Anthropic will be using have not been shared.

In addition to working with Google on the development of its TPUs, last October, Broadcom announced it was working with Anthropic’s biggest rival, OpenAI, to co-develop custom AI hardware.

At the time, the two companies said they would work together on AI accelerators and Ethernet solutions, with deployment targeted to start in the second half of 2026, resulting in 10GW of chips coming online by 2029.

However, since the announcement was first made, no updates on the project have been provided by either company.