Goldman Now Sees $545B in 2027 AI Project Finance, With $320B for Chips
Goldman Sachs' credit team now sees $545 billion in AI project finance for 2027, up from a $300 billion call, with $320 billion of that aimed at chips. Much of the chip debt could come without corporate guarantees, leaving lenders betting on contracts and customers.
- Original source: cryptobriefing.com.
- The source published it on Nov 17, 2026 at 5:15 PM.
- No second outlet has reported it yet. Treat it as a single-source claim.
Goldman Sachs' credit team now expects $545 billion in AI project finance for 2027. Its earlier estimate was $300 billion. The bank splits the money in two: about $225 billion for data center shells — the buildings, power hookups and physical structures — and $320 billion for the chips inside them.
Chip financing is where the curve bends hardest. It has reached $48 billion so far this year, and Goldman expects it to grow almost sevenfold, hitting $320 billion by 2027. In plain terms, Wall Street is preparing to lend more against what sits in the server racks than against the walls around them.
The structure behind it: a special-purpose vehicle owns the chips, keeping the assets and debt off the parent company's balance sheet. A customer signs a lease or service contract for the computing capacity, typically running five years or less, and lenders get repaid from the cash that contract produces. If the customer stops paying, lenders are largely left holding the hardware and the paperwork.
That is the caveat Goldman flags in its research. A large share of the $320 billion in chip financing may arrive without traditional corporate guarantees — no deep-pocketed parent promising to cover the debt if things go sideways. Lenders would be betting on the contract itself and the creditworthiness of the customer who signed it.
Deals are already taking shape. A $40 billion GPU financing package involving SpaceX was recently reported, with Goldman Sachs and Apollo backing the arrangement, and a separate $22 billion chip financing has been tied to Crux AI, a venture connected to Blackstone and Alphabet. Those sit alongside hundreds of billions of dollars in direct investment-grade bond issuance from major hyperscalers expected in 2026.
Goldman's wider AI math has moved too: it now puts 2027 hyperscaler capital expenditure between $1.2 trillion and $1.4 trillion, a top end that would push annualized spending on AI hardware and IP to nearly 4% of US GDP. The research names the pressure points — power supply constraints, regulatory opposition to data center projects, and whether end users can earn enough revenue to meet their contract obligations. The numbers to watch are how much of the $320 billion actually closes, how many deals carry a corporate guarantee, and whether power and permitting snags start delaying the facilities those chips are meant to fill.
TopicsGoldman Sachs · Apollo · Blackstone · Alphabet · SpaceX · Crux AI
Source: cryptobriefing.com
Written by AI from verified reporting.
More stories
- Tech & AI cryptobriefing.com 0 viewsAnthropic IPO order book shuts at 4 p.m. ET, reportedly six times oversubscribed
- Tech & AI breakingthenews.net 0 viewsNvidia Taps Heidrick & Struggles to Run Its CEO Search
- Tech & AI x.com 0 viewsOpenAI launches GPT-6.2 Astra and Sol, billed as its most powerful models yet
- Tech & AI x.com 0 viewsAnthropic eyes IPO price above $80 as demand blows past range

