Anthropic reportedly plans to spend $200 billion with Google Cloud over five years, a move that would account for more than 40% of Google Cloud’s revenue backlog and signals a major shift in cloud-and-AI financing.

Alphabet’s stock rose about 2% after the report, reflecting market optimism about a potential large-scale deal in AI and cloud infrastructure.

Contracts involving Anthropic and OpenAI are said to comprise a significant portion of the $2 trillion cloud-backlog across leading providers like AWS, Azure, and Google Cloud, underscoring the sector’s growth and investment risk.

Note: the article fragment includes generic investing-site risk disclosures and does not provide additional specifics on timelines, scope, or rationale for the reported plan.

Anthropic trains and runs Claude on a mix of hardware, including AWS Trainium, Google TPUs, and Nvidia GPUs, illustrating the hardware diversity behind modern AI systems.

OpenAI is projected to spend about $45 billion on servers this year, up from $17 billion last year, with much of that spending directed to Microsoft, which has invested over $13 billion in OpenAI.

Chipmakers like Nvidia are heavily investing in AI infrastructure, fueling an ecosystem of expensive data-center and chip partnerships.

Anthropic has expanded compute capacity through deals with CoreWeave and almost one gigawatt of capacity via Amazon’s chip initiatives, boosting access to AI hardware for Claude.

Anthropic’s ongoing computing-capacity expansions include CoreWeave agreements and nearly a gigawatt via AWS chips by year-end, reinforcing its hardware access.

Alphabet is moving toward potentially becoming the world’s most valuable company, driven by AI initiatives and a booming cloud business, Supported by rising stock performance.

Reuters could not immediately verify the report; Anthropic did not comment, and Google redirected inquiries to Anthropic, indicating limited independent confirmation.

Despite fueling the AI boom, the surge in spending raises sustainability concerns, including data-center energy use and RAM shortages.