AI co-founder Conor Moore told Benzinga in an exclusive interview. “The banks have gotten drunk on these large IG deals because you don’t have to have a directional view or thesis on the space to be comfortable doing them,” Moore said, referring to investment-grade financing.
When Microsoft Corp. (NASDAQ:MSFT) backs a project, lenders can focus on its ability to pay rather than whether the underlying AI business will succeed.
But the sheer volume of borrowing is testing investors’ appetite. Large bond buyers face limits on their exposure to individual companies, while the flood of AI debt is pushing them to demand higher yields.
” Wall Street’s AI Deals Stretch Lenders Monday’s $63 billion Anthropic financing illustrates the scale of these transactions. Bank of America Corp.
(NYSE:BAC), Citigroup Inc. (NYSE:C) and Morgan Stanley (NYSE:MS) are arranging senior loans for Broadcom Inc.
(NASDAQ:AVGO) AI chips, with Blackstone Inc. (NYSE:BX) providing junior financing.
Several banks considered withdrawing before Broadcom offered better terms, including protection against falling equipment values. Bank of America also struggled to syndicate a $14 billion loan for an Oracle Corp.
(NYSE:ORCL) data center in Michigan and turned to long-term bond investors instead, Karen Fang, the bank’s global head of infrastructure and sustainable finance, told BloombergNEF. She said capital remains available for strong projects.
" Who Finances Smaller AI Companies? AI arranges GPU-backed loans of $20 million to $250 million.
Moore expects average deals to rise from roughly $50 million to $100 million. Smaller operators can deploy GPUs in converted telecom buildings or shipping containers, tapping existing power supplies and potentially avoiding lengthy construction delays.
But their loans are harder for traditional lenders to assess. “The market is starving for standardization,” Moore said.
AI aims to make these loans easier to evaluate, attracting private-credit funds, family offices and eventually regional banks and insurers. On Polymarket, traders put the odds of an AI industry downturn at roughly 10% by December 31 and 23% by June 30, 2027.
The market requires three major distress events within 90 days. Nvidia (NASDAQ:NVDA) also has an interest in expanding access to GPU financing.
In August, Nvidia announced partnerships aimed at mobilizing more than $500 billion in third-party AI infrastructure funding. But lenders remain concerned about GPU depreciation, while financing constraints could slow chip purchases by smaller operators.
AI scores each deal out of 100 on the operator, its customers and the contract itself, rejecting anything below 60.
