Morgan Stanley Says Weaker AI Firms May Turn to Secured Debt

Morgan Stanley strategists Vishwanath Tirupattur and Vishwas Patkar said higher borrowing costs may widen the financing gap in AI infrastructure, leaving the largest hyperscalers better positioned than lower-rated companies to maintain investment.

The four largest hyperscalers have the financial strength to absorb higher funding costs and sustain their plans, the strategists said. Lower-rated AI companies, by contrast, may have to constrain capital spending as credit becomes harder to access.

The financing shift could change how weaker companies borrow. Tirupattur and Patkar expect these firms to rely more on secured, asset-level financing, potentially with credit support from stronger companies, as unsecured borrowing becomes more difficult to obtain.

AI-related debt issuance was nearly $500 billion as of Aug. 5, with hyperscalers accounting for $194 billion in global debt issuance. Those figures describe debt issuance, not total financing needs.

The borrowing divide comes amid a large investment cycle. Broader AI-related investment is projected to exceed $1 trillion globally in 2026, including $581 billion in the United States. Investment in data centers, power systems and chips is projected at $10.3 trillion from 2025 through 2032.