Anthropic explores data center leases, report says
AI startup Anthropic is reportedly pursuing plans to lease and operate its own data centers across the United States, with a combined capacity of over 1 gigawatt. The company has discussed potential financial support from Google to secure these agreements. This expansion aims to meet the growing demand for Anthropic's AI models amidst increasing competition for computing power.
Artificial intelligence startup Anthropic is reportedly advancing plans to lease and operate its own data centers, according to a report by The Information. The developer of the Claude Code AI assistant has signed more than a dozen preliminary agreements for data center leases across the United States, which cumulatively exceed 1 gigawatt in capacity.
Anthropic executives have also engaged in discussions with Google, a current investor in the company, regarding a potential financial guarantee to support these lease payments. Google is also involved in co-designing some of the server chips that could be utilized in these new facilities. Google declined to comment on the report, while Anthropic did not immediately respond to inquiries.
These reported infrastructure expansion plans come as demand for Anthropic's Claude family of AI models continues its growth trajectory. The company has recently focused on increasing its computing capacity as it prepares for an initial public offering in the United States, having confidentially filed for an IPO earlier this month. In April, it was announced that Alphabet, Google's parent company, planned to invest up to $40 billion in Anthropic.
This move by Anthropic underscores the intensifying competition among AI companies for the significant computing power required to develop and operate increasingly sophisticated AI models.