AI's Complex Funding Web: Morgan Stanley Warns of Capital Intensity and Hidden Risks

March 23, 2026
AI's Complex Funding Web: Morgan Stanley Warns of Capital Intensity and Hidden Risks
  • A highly interconnected financing ecosystem has formed in the US AI landscape, with suppliers funding customers and customers supporting suppliers through mechanisms like supplier financing, long-term take-or-pay contracts, revenue-sharing, supplier repurchase agreements, third-party guarantees, IP licensing for access, and equity investments in exchange for compute commitments.

  • Morgan Stanley warns that AI-driven capital expenditures by large-cap peers will run at roughly half of overall capex in the coming years, with capital intensity surpassing the prior Internet era and a tendency toward front-loaded investments that could outpace AI revenue realization.

  • Analysts note these arrangements accelerate infrastructure buildup but risk frontloading demand and shifting risk among counterparties, potentially inflating contracted values and obscuring true leverage on balance sheets.

  • Key players are making sizable commitments: Nvidia’s $30 billion direct OpenAI investment plus GPU leasing and CoreWeave ties; Microsoft’s $13 billion OpenAI stake with further Azure commitments and revenue-sharing; Oracle and AMD’s large orders and partnerships; Amazon’s staged $50-100 billion OpenAI investment and leases; Disney’s $1 billion investment for IP access and a three-year IP license in return for compute access.

  • Many funding agreements cover only a fraction of total compute commitments, with fulfillment contingent on future revenue growth or new funding rounds, underscoring systemic fragility in the money-moving cycle.

  • The piece cautions that while this funding ecosystem enables rapid capacity expansion, investors should stay vigilant to fragility and accounting distortions inherent in these financing cycles.

  • Seven major risks are identified: warrants distorting pricing, off-balance-sheet guarantees hiding true leverage, IP licensing obscuring operating costs, supplier equity investments raising debt risk, high customer concentration increasing counterparty risk, revenue-sharing obscuring true demand, and repurchase agreements inflating reported demand.

  • OpenAI sits at the center of a roughly $100 billion capital flow map, illustrating deep ties with Nvidia, Microsoft, Oracle, AMD, Amazon, Disney, and others.

Summary based on 1 source


Get a daily email with more AI stories

More Stories