The AI debt dilemma hits Big Tech earnings (www.axios.com)

Big Tech is increasingly using debt to finance a costly AI buildout, and Wall Street is starting to panic. The market reaction was visible when Meta fell 11% after reporting strong results but raising capital-spending plans, underscoring investor sensitivity to rising AI-related capex. Analysts note the nuance: companies with strong balance sheets and clear AI demand (e.g., Microsoft) can signal growth by spending more, but others may be borrowing to chase uncertain future revenue. Morgan Stanley projects Oracle’s debt could double to over $290 billion by fiscal 2028, and specialty players like CoreWeave and Crusoe are also taking on leveraged positions. The technical implications are twofold. First, AI hardware cycles — every new Nvidia chip generation — force continual, expensive upgrades, lengthening and enlarging capex commitments. Second, some financing is moving off-balance-sheet via private loans or structured deals, making true leverage harder for investors to assess. If AI investments fail to yield productivity gains or revenue, heavy debt loads could create wider market stress reminiscent of past bubbles (housing, dot-com). The key things to watch: whether AI drives measurable revenue/productivity to justify ongoing upgrades, the mix of debt vs. equity financing, and the transparency of off-balance-sheet obligations — all factors that will determine whether this is disciplined investment or a speculative bubble.