Many new hyperscaler bonds trade wider within six months of issuance, BondCliQ CEO Chris White told CNBC’s Fast Money. He says the artificial intelligence (AI) debt glut may also be lifting US government borrowing costs.

White, whose firm BondCliQ supplies corporate bond pricing data, sees no major default risk. However, he expects lenders to demand higher yields from hyperscalers, the tech giants building AI data centers.

Why Are New AI Bonds Losing Ground So Soon?

Bond investors react over weeks and months, White told CNBC, unlike stock traders who move day to day. He pointed to CoreWeave, an AI cloud computing provider, and Oracle as cases where bond buyers doubt the long-term story.

Meanwhile, Apollo Global chief economist Torsten Slok has tied rising default-insurance costs on hyperscaler debt to weakening credit fundamentals.

A wider spread means investors demand more yield over safer benchmarks. White called that pattern abnormal for large institutional buyers.

Therefore, he expects those buyers to price in that history when hyperscalers return to borrow.

“investors will not be fooled twice.”

Chris White, CEO of BondCliQ, on CNBC

White put CoreWeave at about 5% of the $258 billion in AI-linked debt sold over the past year. It must refinance within five to seven years, leaving it exposed to higher short-term rates.

In contrast, Oracle’s debt runs for decades, so rate rises pose less refinancing pressure.

Could the AI Debt Glut Raise Borrowing Costs for Everyone?

Microsoft is the only hyperscaler with a top AAA credit rating, White noted. Its bond pricing therefore sets the benchmark for peers. If Microsoft offered sweeter terms, he said, rivals would be expected to match.

He linked the oversupply to US Treasuries, citing White House concern about long-term yields and direct bond-buying programs. He framed the connection as his own view.

That matters for equities, since Ray Dalio has warned of a shrinking cushion for stocks as yields rise.

White expects lenders to demand higher yields from every borrower, from other companies to Washington itself. If he is right, AI funding costs could help set the price of credit well beyond tech.