AI Infrastructure Spending Hits Record $800 Billion Amid Bubble Warnings
Financial analysts and industry leaders warn of an AI bubble as hyperscaler capital expenditure nears $800 billion against lagging revenue growth.
Industry leaders and financial analysts are warning of a potential bubble in the artificial intelligence sector as infrastructure spending reaches record levels. Hyperscalers' capital expenditure is projected to reach nearly $800 billion in FY2026, pushing total liabilities above $4 trillion.
Broadcom Inc CEO compared some AI investors to "casino gamblers" and warned that financing this infrastructure could become a major challenge. These concerns are highlighted by revenue gaps at major firms; OpenAI reported annualized revenue of $50 billion by late September, falling $20 billion short of the $70 billion expectation. Additionally, investor Ray Dalio warned that rising interest rates and massive debt could cause the bubble to burst.
Conversely, some market participants argue the spending is backed by genuine demand. Microsoft Azure has surpassed $100 billion in annual revenue, and Amazon Web Services is reporting its fastest growth in years. While firms like Anthropic remain unprofitable as they push for a valuation exceeding $2 trillion in an upcoming IPO, proponents of the sector point to these strong cloud orders as evidence that the buildout is justified.
