OpenAI and Oracle have announced plans to build an additional 4.5 gigawatts of AI-focused data centre infrastructure, extending their collaboration under the ambitious Stargate initiative. The new phase pushes the total development target past 5 GW, reinforcing the project’s aim to give the United States an edge in AI infrastructure scale and performance.
While exact locations and funding breakdowns remain undisclosed, the expansion builds on the initial commitment made in January 2025, when the Stargate programme was launched at the White House. At that time, the initiative—backed by Microsoft, SoftBank, and other partners—outlined an investment roadmap of up to $500 billion to support large-scale AI deployments.
The upcoming facilities will reportedly operate on over 2 million AI-optimized chips and are designed to support workloads from generative AI tools like ChatGPT and Copilot. Analysts say the new capacity is essential to meet escalating demand in sectors ranging from healthcare and finance to defence and education.
Funding questions persist as stakeholders adjust scope and pace
Despite the scale of the announcement, doubts remain around funding feasibility. Reports earlier this year suggested that OpenAI and SoftBank were expected to contribute $19 billion each toward Stargate, but recent updates hint at scaled-back deployment goals. The Wall Street Journal noted that the first phase may now result in a single facility, potentially located in Ohio, by the end of 2025.
Oracle has not responded to recent media queries, while the White House has declined to comment on the latest announcement. Nonetheless, industry observers view the expansion as a key indicator of sustained momentum behind U.S.-led AI infrastructure development.
With AI increasingly shaping economic and geopolitical competition, initiatives like Stargate signal an evolving landscape where data centre capacity is as strategic as software innovation itself.
