Anthropic's Billions in Microsoft Commitments Add to Its Mounting Compute Bill

Anthropic's IPO is shaping up to be the ultimate Rorschach Test for the AI bull market.

Bulls see one of the most revolutionary companies in history, which, just five years after its founding, is leading the generative AI push, and whose Claude chatbot and related tools have become essential for millions of businesses and everyday users.

Bears, on the other hand, see a company writing checks that its chatbots can't cash. It has over $500 billion in future spending commitments and reported an $80.7 billion generally accepted accounting principles (GAAP) loss in the first half of the year, though that was mostly related to changes in valuations of its convertible debt due to the company's soaring valuation.

The bull vs. bear debate is about to get real once the AI stock goes public at an expected valuation of $2 trillion, but investors got a chunk of information to chew on from its S-1 prospectus out on Monday, and one of the noteworthy factoids was Anthropic's relationship with Microsoft (NASDAQ: MSFT).

Image source: Getty Images.

Anthropic will pay Microsoft at least $45 billion

Microsoft was the last of the three big cloud computing platforms to invest in Anthropic, in part due to its close relationship with OpenAI, but the Windows parent is one of several cloud platforms Anthropic has contracted with for computing power.

According to the S-1, Anthropic will spend at least $44.9 billion between November 2026 and May 2033, regardless of its usage, and the agreement is mostly non-cancellable.

Anthropic has committed to spending $114.5 billion during a similar time frame with Alphabet, $110 billion with Amazon through 2036, and approximately $3.1 billion per month with Space Exploration Technologies (NASDAQ: SPCX) through 2029.

That list is not exhaustive of its spending commitments and implies at least $400 billion in cloud infrastructure spending over the next several years.

What it means for Anthropic

Anthropic spent nearly $9 billion on compute and infrastructure in the first half of the year, and that was its biggest line item cost by far.

The company's future cloud spending commitments go hand in hand with its growth. Computing power is what drives the business, and spending on it is as elemental as a restaurant spending on food.

The numbers may be eye-popping, and they do imply a lot of risks in the business if demand doesn't show up, but computing power is scarce right now, and Anthropic needs to make such moves to secure compute, or it will struggle to grow.

Investors scared off by such commitments may be better off taking the other side of the bet and investing in the cloud computing giants. Of course, those companies are spending hundreds of billions on building out those data centers to provide companies like Anthropic with the computing they need, and stocks like Microsoft are clearly less risky than AI start-ups like Anthropic.

The bull vs. bear debate will continue no matter what happens when Anthropic goes public, but the AI lab's debut will give investors a clear idea of where AI sentiment stands today. There's a lot of upside and downside to Anthropic stock, especially in the coming months.

Jeremy Bowman has no position in any of the stocks mentioned. The Motley Fool has positions in and recommends Microsoft. The Motley Fool has a disclosure policy.