Oracle, OpenAI Halt Texas Data Center Expansion * Science and Technology Times

Oracle and OpenAI have ended plans to expand a major AI data center in Texas after negotiations stalled.
The companies will shift the planned capacity to other sites while continuing broader infrastructure development.
Meta has reportedly explored taking over the abandoned expansion area as competition for AI‑ready facilities intensifies.

Stalled Negotiations Shift Expansion Plans

Oracle and OpenAI have dropped their plans to expand a flagship artificial intelligence data center in Texas, according to reports from Bloomberg News. Discussions reportedly slowed due to disagreements over financing and evolving requirements from OpenAI, ultimately leading the companies to abandon the proposed addition. The expansion was originally part of the large‑scale Stargate initiative, a project involving up to $500 billion in investment and 10 gigawatts of planned capacity. Announced in January 2025, the initiative includes contributions from SoftBank Group, OpenAI and Oracle.

The companies had previously outlined a potential 600‑megawatt expansion near the main Stargate site in Abilene, Texas. That additional capacity will now be allocated to one of several other data center campuses currently under development. A source familiar with the matter told Reuters that the Abilene location remains active, with eight buildings planned and two already operational under Oracle Cloud Infrastructure. The broader plan to develop 4.5 gigawatts of new data center capacity continues unchanged.

Technology firms have been investing heavily in data centers to support generative AI services such as ChatGPT and Copilot. These systems require substantial computing power, driving demand for large‑scale infrastructure capable of supporting advanced AI workloads. The rapid expansion of AI facilities has intensified competition for suitable sites, energy resources and semiconductor supply. As a result, shifts in development plans can create opportunities for other companies seeking additional capacity.

The decision to halt the Texas expansion has already drawn interest from other industry players. Bloomberg reports that Meta Platforms has explored leasing the planned expansion site from developer Crusoe. Nvidia reportedly helped facilitate the discussions, reflecting the company’s role in shaping the AI hardware ecosystem. The involvement of multiple major firms highlights the strategic importance of securing reliable data center space for future AI growth.

Stargate Initiative Continues Despite Setback

The Stargate initiative remains one of the most ambitious AI infrastructure projects announced to date. Its goal is to support the massive computational requirements expected from next‑generation AI models and services. Although the Texas expansion has been canceled, the overall project continues to move forward across multiple locations. Oracle and OpenAI still plan to develop several gigawatts of additional capacity as part of the broader effort.

The Abilene site remains a central component of the initiative. Oracle Cloud Infrastructure is responsible for operating the facility, which is designed to support high‑density AI workloads. Two of the eight planned buildings are already active, providing early capacity for ongoing AI development. Additional buildings will come online as construction progresses, contributing to the long‑term goals of the Stargate project.

Nvidia’s involvement underscores the importance of semiconductor supply in large‑scale AI deployments. The company’s GPUs are widely used for training and running advanced AI models, making them essential for facilities like Stargate. According to Bloomberg, Nvidia intervened to ensure that its chips would be used in the expansion rather than those from competitor AMD. This reflects the competitive dynamics within the AI hardware market, where access to cutting‑edge components can influence project decisions.

The shift in expansion plans does not appear to affect the companies’ broader collaboration. Oracle and OpenAI continue to rely on Nvidia hardware for their AI infrastructure, and the remaining phases of the Stargate initiative remain on schedule. The decision to relocate the 600‑megawatt expansion may streamline development by aligning capacity with evolving operational needs. It also demonstrates the flexibility required to manage projects of this scale.

Industry Competition Shapes Data Center Strategy

The abandoned expansion highlights the growing competition among technology companies for data center resources. AI development has accelerated demand for power, land and specialized hardware, creating pressure on companies to secure long‑term infrastructure. Meta’s interest in the Abilene site illustrates how quickly opportunities can shift in this environment. Companies that can move rapidly may gain advantages in securing strategic locations.

Negotiations between Oracle and OpenAI reportedly slowed due to financing challenges and shifting requirements. These factors are common in large‑scale infrastructure projects, particularly those involving emerging technologies. As AI workloads evolve, companies must frequently reassess their capacity needs and deployment strategies. This can lead to changes in project scope, location or partnerships.

The broader industry trend shows no signs of slowing. Major firms continue to invest billions of dollars into new data centers to support AI services. These facilities require significant energy resources, prompting companies to explore new power agreements and alternative energy sources. The scale of these investments reflects the central role AI is expected to play in future digital services.

Meta, Oracle, OpenAI and Nvidia all declined to comment or did not respond to requests for comment regarding the latest developments. Their silence suggests that negotiations and strategic planning may still be ongoing behind the scenes. As the AI infrastructure landscape continues to evolve, further adjustments to major projects like Stargate are likely.