Blackstone’s AI-credit argument centres on customers behind the infrastructure
Jas Khaira’s defence of data-centre financing points to counterparty quality, while an earlier Firmus facility illustrates the scale of the firm’s exposure.
Blackstone’s Jas Khaira argues that highly creditworthy customers underpin much AI-infrastructure lending, weakening comparisons with the 2008 banking crisis.
The argument directs attention to the revenue arrangements behind a facility, rather than judging the credit solely from the amount of computing equipment being built. It remains an investor’s assessment, not proof that a financing cannot lose money.
An earlier deal shows the scale
Firmus announced a $10 billion debt financing facility on February 9, led by Blackstone-managed funds and supported by Coatue. The company said it would fund the next phase of Project Southgate, its Australian AI Factory rollout based on NVIDIA’s DSX reference architecture.
That is an earlier financing announcement, not a new October transaction. A facility’s announced size should also be distinguished from the amount drawn, the capacity installed and revenue earned from it.
A contract and the asset answer different questions
A customer agreement may support a lender’s view of future payments, while equipment can provide collateral. Each depends on specific terms: who is obliged to pay, when obligations begin and what happens if equipment delivery or customer demand changes.
A strong counterparty does not automatically remove construction risk, timing risk or uncertainty about the asset’s future value. Conversely, a large borrowing figure does not by itself establish an impending default.
The useful part of the debate is that it makes those distinctions explicit. AI infrastructure links financial contracts with a physical delivery programme and a service that customers must ultimately use.
Khaira’s position adds a lender’s perspective to that discussion. Realised credit outcomes will depend on the transactions and operating performance, rather than the analogy used to describe the cycle.
Sources
- [Khaira’s AI-credit view](https://techcrunch.com/2026/10/21/blackstones-jas-khaira-says-fears-over-ai-buildout-debt-are-overblown/) — TechCrunch - [Firmus announces $10 billion financing led by Blackstone and Coatue](https://firmus.co/newsroom/firmus-secures-us10-billion-financing-led-by-blackstone-and-coatue-to-scale-energy-efficient-ai-infrastructure) — Firmus