U.S. artificial intelligence startup Anthropic has pledged to spend $200 billion on Google Cloud services and chips over the next five years as part of a recent deal, people familiar with the matter said. The commitment means Anthropic accounts for more than 40% of the backlog revenue Google disclosed to investors last week. These unconfirmed orders reflect contractual commitments from cloud service customers.
Boosted by the news, the stock price of Google parent company Alphabet rose about 2% in after-hours trading on Tuesday.
Anthropic signed an agreement with Google and its chip partner Broadcom in April this year to obtain several gigawatts of Tensor Processor Unit (TPU) production capacity, which is expected to be put into use starting in 2027.
Alphabet will also invest up to $40 billion in Anthropic to further deepen its cooperation with the artificial intelligence startup. Anthropic is also a competitor of Alphabet in the global artificial intelligence race.
According to reports, contracts involving Anthropic and OpenAI now account for more than half of the $2 trillion in unconfirmed orders from major cloud providers such as Amazon Web Services, Microsoft Azure, and Google Cloud Platform.
Due to strong market demand for its Claude series of artificial intelligence models, Anthropic recently signed a series of important agreements to obtain more computing power.
Last month, Anthropic struck a multi-year deal with cloud infrastructure company CoreWeave and plans to acquire nearly 1 gigawatt of computing power via Amazon chips by the end of the year.
Anthropic said it trains and runs Claude models on a variety of artificial intelligence hardware, including Amazon Web Services’ (AWS) Trainium, Google TPU and Nvidia GPU.
Meanwhile, Alphabet is on track to overtake Nvidia to become the world’s most valuable company on the back of a record share price surge driven by its artificial intelligence and booming cloud computing business.