OpenAI Commits $1.4 Trillion to Data Center Expansion, Targeting 30 Gigawatts Capacity

San Francisco, CA – OpenAI CEO Sam Altman recently announced a staggering commitment of approximately $1.4 trillion towards infrastructure development, aiming to secure 30 gigawatts (GW) of data center capacity. This ambitious plan, unveiled during a livestream on October 28, 2025, follows a significant restructuring agreement with Microsoft and underscores the company's aggressive pursuit of advanced AI capabilities. The announcement has sparked discussions regarding the financial feasibility and the complex web of obligations underlying such a massive undertaking.

The $1.4 trillion investment is intended to be deployed over the next few years and clarifies numerous prior announcements concerning partnerships with major players like AMD, Broadcom, NVIDIA, and Oracle. Altman also outlined a long-term aspiration to achieve a build rate of 1 GW of new capacity per week, with an estimated cost of $20 billion per GW. This unprecedented scale is deemed essential to meet the escalating computational demands of future AI models.

However, industry experts and analysts have raised questions about the financial and logistical viability of these commitments. Reports suggest that only a fraction of the targeted 30 GW, approximately 4.5 GW, is currently backed by confirmed financing, such as a $30 billion partnership with Oracle. The remaining capacity lacks secured funding, power agreements, or necessary regulatory approvals. OpenAI, despite generating substantial revenue—$4.3 billion in the first half of 2025—also reported a net loss of $4.7 billion during the same period, burning billions to fuel its rapid expansion.

The financial strategy involves a unique "vendor financing" model, where partners like Nvidia invest in OpenAI, which then uses that capital to purchase their chips. AMD has also granted warrants for equity in exchange for deployment agreements, and Oracle is constructing data centers specifically for OpenAI's workloads. This approach, while aligning interests, concentrates risk and creates a complex structure of obligations. Regarding the seniority of these spend obligations, as queried by Hersh Desai on social media, "Who gets paid first if OpenAI is cash-crunched and capacity comes online at the same time?", the intricate nature of these vendor-financed deals and traditional debt structures would likely prioritize secured creditors over equity holders in a distressed scenario.

To sustain these colossal investments, OpenAI aims to significantly boost its annual revenue into the "hundreds of billions," with an initial public offering (IPO) considered a probable future path. The company's audacious infrastructure drive highlights the intense competition and capital requirements in the race to develop superintelligence, pushing the boundaries of traditional corporate finance and market expectations.