International AI, a U.S. firm that builds single-tenant, air-gapped information facilities for governments and enterprises, has closed $441 million in financing, its first-ever debt increase. The senior secured credit score facility was led and organized by J.P. Morgan, with participation from a gaggle of lenders the corporate didn’t title.
The announcement lands the identical week one other GPU infrastructure operator, Nscale, lined up a $900 million credit score facility from a lender group that additionally included J.P. Morgan — two nine-figure debt packages for non-hyperscaler AI capability inside days of one another.
Co-Founder, Director, and CEO Sami Issa tied the increase to contracted demand. “We’re considerably accelerating our capital-raising efforts to fulfill the distinctive demand we’re seeing for devoted, single-tenant, air-gapped AI infrastructure,” Issa stated. “With $6.2 billion in contracted revenues, together with $1 billion in contracted revenues already constructed and delivered to clients, we’re scaling quickly to fulfill this demand.”
What International AI is definitely constructing
Based in 2024, the corporate describes itself because the world’s first sovereign AI hyperscaler: it designs, builds, powers, and operates information facilities the place one buyer occupies all the atmosphere, bodily remoted from different tenants and from public networks. Its mannequin spans land, vitality, building, liquid cooling, and GPU-dense compute, geared toward nations and enterprises that need AI coaching and inference inside their very own jurisdiction quite than on shared public cloud.
The {hardware} program is NVIDIA-based. The corporate, an NVIDIA Cloud Companion, owns and operates an information middle in Endicott, New York, working GB200 NVL72 clusters. On December 3, 2025, it introduced a partnership with HUMAIN, the AI firm backed by Saudi Arabia’s Public Funding Fund, to develop large-scale AI information facilities in america; that launch additionally disclosed that NVIDIA had joined a current International AI fundraising spherical.
On March 16, 2026, the corporate stated it had accomplished a GB300 NVL72 deployment it calls the most important such cluster in New York State, is actively deploying 7,000 GB300 programs, and plans to roll out NVIDIA’s next-generation Vera Rubin NVL72 racks throughout its U.S. footprint.
The capability roadmap behind the increase
The debt comes towards a buildout the corporate has already scheduled publicly: 16 megawatts of preliminary crucial IT load at Endicott, ramping to 100 megawatts by means of 2026, with 250 megawatts deliberate for 2027 and 1 gigawatt of crucial capability focused by the tip of 2029.
For a corporation based in 2024, the $441 million senior secured facility towards $6.2 billion in contracted income is its first-ever debt increase, following earlier fundraising rounds that included NVIDIA. The said subsequent milestone is the ramp to 100 megawatts at Endicott by the tip of 2026, adopted by the Vera Rubin NVL72 deployments and the 250-megawatt mark in 2027.
