AI infrastructure companies that borrowed heavily to build data centers are facing fresh financial strain as bond yields spike. CoreWeave, which went public last year, disclosed in its latest quarterly filing that a 1-percentage-point rise in interest rates would boost its annual interest expense by $30 million, based on its outstanding floating-rate debt. The warning highlights how sensitive the sector has become to the cost of capital.
Oracle provided an early flashpoint this week when its stock fell after Bloomberg reported the company had sent a force majeure notice tied to its New Mexico data center campus, Project Jupiter. The notice seeks to delay payment if the project does not come online as expected in 2028. Oracle responded that the project "remains on our planned schedule," but the episode underscores how rising rates are complicating large-scale builds.
Beyond financing costs, AI companies face mounting political headwinds. A recent NBC News Decision Desk Poll found 69% of respondents oppose data center construction in their local area, and Texas Governor Greg Abbott has temporarily halted data center-related environmental permits. Meanwhile, executives at Anthropic and OpenAI have urged a slower pace of AI development after researchers raised concerns about advanced models.
Despite these pressures, demand for AI services shows no sign of slowing. Meta's Muse personal assistant app surpassed 2.5 million downloads in its first two weeks, briefly topping ChatGPT on Apple's App Store, and Evercore's Mark Mahaney projects it could reach 100 million users within six to 12 months. The tension between surging demand and rising borrowing costs is likely to define the next phase of the AI buildout.