In its second-quarter earnings report on Wednesday, tech giant Microsoft reported $37.5 billion in capital expenditures, exceeding market estimates by more than a billion. The spending was up 66% from a year earlier, and roughly two-thirds of it was primarily spent on GPUs and CPUs, Microsoft executives said in an investor call.
A few months ago, a report like this would have sent (and did send) Microsoft stock soaring. But on Wednesday, it had the opposite effect, and the stock went down 7%.
As worries over an AI bubble simmer, the market is more desperate than ever to see tangible revenue returns that can reignite belief in the great financial promises of the technology, rather than just another huge spending commitment.
But accompanying Microsoft’s record spending was slowing cloud growth.
Revenue from Microsoft’s cloud services grew by 39% this quarter, down from 40% growth in the first quarter. During the investor call, Microsoft CFO Amy Hood attributed this discrepancy between capital expenditure and cloud growth at least partially to Microsoft allocating GPUs and cloud capacity to internal teams as well. Customer demand for cloud is still outpacing supply, Hood said.
But even if the slowing cloud growth can be explained away, what also got investors anxious was Microsoft’s reliance on AI giant OpenAI. 45% of Microsoft’s remaining cloud commitments are solely from OpenAI.
Although OpenAI used to be the silver bullet for finance, growing uncertainty over the startup’s road to profitability and the risks associated with its ability to pay for towering, ambitious dealmaking has made some start to view any dependence on the AI darling as a potential burden.
OpenAI has signed trillions of dollars worth of deals this past year, despite its $20 billion annualized revenue. Lately, the market has started questioning these overcommitments, as concerns over a potential AI bubble mount.
If it continues to take too long for AI investments to start translating to actual gains or if somehow it turns out OpenAI cannot pay for its piling commitments, it could lead to a sharp market correction and spell trouble for the U.S. economy, which it seems is currently held up by AI investment.