Is the AI race threatening tech giants?

“Al Arabiya” – A recent report revealed that the artificial intelligence boom driven by major technology companies relies on off-balance-sheet financial commitments totaling approximately $3.1 trillion, a figure roughly five times higher than the annual capital expenditure of the largest cloud computing firms, raising growing questions about the sector’s ability to self-finance this expansion.

Market movements showed that investors have not yet panicked following the release of these figures. Alphabet’s share price fell by about 1 percent since August 16, while Amazon lost 2 percent and Microsoft dropped 3 percent. Meta’s stock declined by 8 percent, a drop partly linked to a case concerning child safety on social media platforms.

According to Bloomberg, investors focused on four key elements during recent earnings reports: strong demand, accelerating AI-related revenue, profitability and cash flow sustainability, and the ability of executive teams to demonstrate that massive infrastructure spending will translate into future returns.

CoreWeave underscored the importance of these factors after its share price surged 18 percent following the announcement of its results on August 11. The rally was driven by a 112 percent revenue growth to $3.45 billion, an upward revision of its 2026 outlook, a backlog of orders worth $104 billion, and new customer commitments exceeding $25 billion.

However, pressure is already mounting on the cash flows of major technology companies. Alphabet’s free cash flow turned negative in the second quarter of 2026 for the first time since 2004, while Amazon also recorded negative free cash flows, with expectations that this trend will continue through the end of the year.

The $3.1 trillion figure comprises approximately $1.2 trillion in inactive lease contracts and $1.9 trillion in future purchase commitments, according to disclosures from companies including Alphabet, Amazon, Meta, Microsoft, Oracle, Nvidia, Broadcom, AMD, and SpaceX. The majority of these funds are allocated to building AI infrastructure.

These companies have increasingly relied on specialized financing structures and special-purpose vehicles to keep these commitments off their balance sheets. This approach helps them maintain high credit ratings and preserve the liquidity needed for investment, share buybacks, and dividend distributions.

These practices differ from the model associated with Enron before its collapse in 2001. Reports indicate that technology companies disclose these arrangements to investors, although estimating their future impact on cash flows remains complex due to differences in accounting assumptions used.

Massive investments have already strained liquidity. According to Amazon’s disclosures, free cash flow over the past 12 months dropped to $1.2 billion, down from $25.9 billion previously, before turning into a deficit of $7.6 billion by the second quarter of 2026 due to intensive spending on AI infrastructure. Meta’s free cash flow also fell from over $43 billion to levels nearing single digits.

Risks are increasing if demand for AI services slows down, as many current commitments are non-cancelable or based on “take-or-pay” contracts. In such a scenario, expenses would remain fixed while the revenues capable of covering them would decline.

Oracle is viewed as the most exposed to risk among major AI infrastructure providers, with a significant portion of its expected growth dependent on OpenAI, alongside projected negative free cash flows of $23.7 billion in 2026.

In contrast, Microsoft and Meta appear better positioned thanks to diversified revenue streams from software and advertising, while emerging cloud computing firms such as CoreWeave, Nebius, and OpenAI itself remain highly sensitive to any slowdown in demand or tightening of financing conditions.

Despite these concerns, some analysts remain optimistic. Raj Joshi of Moody’s notes that tech giants only build new data centers after securing clear commitments or contracts from customers, pointing out that the sector still faces a shortage of operational capacity relative to the growing demand for training models and running AI applications.

While markets do not currently appear concerned, any decline in demand for AI services could turn these massive commitments from a growth driver into a burden, pressuring cash flows and stock valuations for companies such as Microsoft, Amazon, Google, Meta, and Oracle.