SAN FRANCISCO — Meta Platforms reported a 91 percent drop in free cash flow last quarter, the steepest decline among major technology firms racing to fund artificial intelligence infrastructure.
Amazon, Alphabet and Tesla posted negative cash flow in the same period. The reversals reflect accelerating capital expenditures on data centers and AI computing hardware, including Nvidia's H100 and B200 GPUs and high-bandwidth memory—components essential for training and deploying large language models.
Demand for those components has pushed procurement costs sharply higher. High-bandwidth memory prices have surged alongside GPU costs, compressing near-term cash generation even as revenue grows.
CEO Mark Zuckerberg has made AI the company's top engineering priority. Meta plans to spend tens of billions on AI infrastructure over the next several years. Amazon Web Services, Alphabet's Google Cloud and Microsoft Azure are executing similar capital-intensive buildouts to capture market share in cloud AI services.
Sustained negative cash flow strains balance sheets. While all four companies hold substantial cash reserves, continued capital burn limits share buybacks and dividends—a meaningful shift from the free cash flow abundance that defined the prior decade.
The capital required to compete at the AI frontier creates a barrier to entry that favors the largest players. That dynamic consolidates power among the few firms with the financial capacity to fund these buildouts, while smaller AI startups remain dependent on venture capital and access to hyperscaler cloud platforms.
Investors appear unbothered for now. Meta rose 3.3 percent to $556.71. Amazon climbed 15.3 percent to $271.58 and Alphabet gained 6.7 percent to $356.13. Tesla added 0.8 percent to $311.21.
The core wager is that AI investments will generate new revenue streams and operational efficiencies. The risk is uncertain return on investment and rapid hardware obsolescence in a market where AI model capabilities and pricing remain in flux.
