Escalating Costs Amidst Cash Flow Decline
On July 29, Meta raised its 2026 capital expenditure guidance to a range of $130 billion to $145 billion. This move occurred against the backdrop of a 91% year-over-year decline in second-quarter free cash flow, which fell to $784 million. This stark contrast between massive spending and shrinking liquidity has intensified investor skepticism regarding the long-term payoff of the company's aggressive AI buildout.
The Strategic Rationale
Chief Financial Officer Susan Li has defended the spending, arguing that the industry has historically underinvested in AI compute. Meta management asserts that existing capacity is extremely valuable and will remain tight for the foreseeable future. The company plans to deploy up to 14 gigawatts of compute next year, aiming to monetize this infrastructure by selling access to hosted models or renting out spare machine time, effectively pivoting from a purely ad-centric business model to a hybrid cloud-and-AI vendor.
The Wall Street Test
The company's thesis rests on two unproven bets: that demand will outstrip supply for several years and that Meta can successfully transition into enterprise-grade compute sales. Unlike Amazon and Microsoft, which have established enterprise sales channels and long-standing cloud ecosystems, Meta lacks a comparable footprint. Analysts have warned that if the company fails to demonstrate concrete revenue from these new initiatives or show signs of improving cash generation, it could face negative free cash flow by the latter half of 2026—a scenario not seen since the company's IPO. As the market signals growing impatience, Meta faces a critical test to prove that its high-risk, debt-fueled capital expenditure will eventually translate into durable margins.
Source: AI Daily News. This story was rewritten and translated for our readers.