KOHb - Getty Images

Mixed views on Stargate AI infrastructure plan with OpenAI

The OpenAI-Oracle-Softbank collaboration highlights the determination of the U.S. to maintain its lead in AI technology. However, some question OpenAI's lead and involvement.

The new data center-based U.S. infrastructure project for AI technology introduced by President Trump and OpenAI was met with enthusiasm and criticism, including concern over the plan's massive compute demands and cost.

Newly inaugurated President Donald Trump on Tuesday revealed that the Stargate Project will invest up to $500 billion in the next four years to build AI infrastructure to support AI development.

OpenAI, Oracle, Middle East AI venture capital fund MGX and SoftBank are the initial equity investors in Stargate. The project is expected to create more than 100,000 jobs in the country, according to OpenAI.

Microsoft, Nvidia and U.K. chipmaker Arm will also be part of the project, with the key buildout underway in Texas. It's unclear why Texas is the first state involved with the project, and if other states will be included.

Advancing the U.S. lead in AI

Stargate addresses what tech vendors and the administration see as a critical need to establish and maintain the U.S as the leader in AI technology.

"It reflects a public-private partnership to accelerate AI development," said Lydia Clougherty Jones, an analyst at Gartner. "The missing piece is the energy that's needed to develop AI at the higher levels, which is desired by the administration."

The project's funding will go toward building data centers and energy infrastructure so the U.S. can maintain its global leadership status while overcoming the huge energy demands of developing and using AI and generative AI technology.

Energy demand is a big problem as AI technology continues to expand and grow, said Mel Morris, CEO of Corpora.ai, developer of an AI research engine.

Both administrations have had the ambitions to develop these data centers with the requisite energy infrastructure in the United States.
Lydia Clougherty JonesAnalyst, Gartner

"The elephant in the room on this one is power generation, and therefore we have to start to consider how we're going to actually provide the levels of electrical power these things are going to need in order to fuel that development," Morris said.

Trump is not the only president who has seen AI infrastructure and energy consumption as a critical challenge.

Former president Joe Biden on Jan. 14 signed an executive order to provide support for the massive need for fast-growing AI data centers.

"Both administrations have had the ambitions to develop these data centers with the requisite energy infrastructure in the United States," Clougherty Jones said.

Mixed reactions

The new project has received mixed reaction, including unexpected criticism from xAI CEO Elon Musk, a key backer of President Trump.

In response to OpenAI's post about Stargate on X, Musk wrote: "They don't actually have the money."

While Altman first responded to Musk by calling him "the most inspiring entrepreneur of our time," he later followed up by asking Musk to put the country first in his new role as head of the Department of Government Efficiency, or DOGE.

Musk was joined by others in critiquing the new venture.

Tyler Saltsman, a former U.S. Army officer and founder of generative AI startup EdgeRunner AI, said he sees Stargate as an opportunity for the U.S.. However, he questioned OpenAI's lead role and motive.

"It's a great signal that the administration is prioritizing investments for AI for the U.S," Saltsman said.

He added that the biases in OpenAI's training data are not clear, which is problematic and a threat to national security.

"As we build AI that becomes the backbone of all of our infrastructure, it needs to think like an American. It needs to have American values, and it needs to be culturally aware to Americans to serve Americans," Saltsman continued. "We need to understand that training data, where they're getting the training data and what it looks like because if we don't, it becomes a black box."

Meanwhile, OpenAI's recent internal changes are also a cause for concern, said Davi Ottenheimer, vice president of trust and digital ethics at private data storage firm Inrupt. Ottenheimer was referring to most of OpenAI's founding board members having left the company.

"I wouldn't look to OpenAI for innovation," Ottenheimer said. "I saw a big shift in the board, and I think that was a pretty clear warning sign that there's something rotten in OpenAI."

Softbank and Oracle are also troubled, he maintained. Oracle is known for being behind in technology, he continued.

"Some of the most troubled organizations in the world are being built together to form a partnership that's for Texas," he continued. "It's for AI in Texas so, that doesn't sound right. This looks like a lot of moving money around and looking like we're doing something."

Government regulation of Stargate is critical to ensure that the project's creators will deliver what they have promised, he continued.

Moreover, there are numerous moving parts to the project, including the different alliances involved, said Michael Bennett, associate vice chancellor for data science and AI strategy at the University of Illinois Chicago.

"This is a complicated mix of political, geopolitical and financial relationships that could make this a complicated project going forward," Bennett said. By geopolitical, Bennett references the global competition between the U.S., China and other technologically advanced nations to exploit AI.

Whether Stargate fulfills its promise will be seen in four years. In the meantime, the project itself might face some challenges depending on how it's rolled out in other states. "We have to imagine the communities in which these new centers are proposed will at least in some instances have something to say about the energy requirements and the proposed solution to providing energy," Bennett said.

Esther Shittu is an Informa TechTarget news writer and podcast host covering artificial intelligence software and systems.

Dig Deeper on AI infrastructure