OpenAI's $500 billion Stargate infrastructure initiative is considering Utah as a primary deployment location.

Artificial intelligence developer OpenAI on Thursday announced its list of 16 states, including Utah, that could potentially play host to nodes on a network of massive new data centers, the computing backbone for the privately funded $500 billion Stargate Project announced last month. At a Thursday press briefing, OpenAI detailed plans to build an interconnected system of five to 10 data center campuses, each roughly 1GW in scale that, combined, would support OpenAI's "frontier AI models." While the first Stargate data center campus is already under construction near Abilene, Texas, Utah is on the list of future data center buildout sites "under active consideration" by OpenAI. Other states competing for the investments include Arizona, California, Florida, Louisiana, Maryland, Nevada, New York, Ohio, Oregon, Pennsylvania, Virginia, Washington, Wisconsin and West Virginia.

Stargate brings together some of the biggest global players in tech. OpenAI CEO Sam Altman, Oracle founder Larry Ellison and SoftBank CEO Masayoshi Son gathered with President Donald Trump at a Jan. 22 White House press conference to announce the project, a privately funded effort to invest $500 billion in artificial intelligence infrastructure, including massive data centers and new energy systems over the next four years. According to OpenAI's announcement, MGX, a $100 billion investment arm of the United Arab Emirates sovereign wealth fund, is an equity partner in Stargate along with OpenAI, Oracle and SoftBank. Other partners include Microsoft, a major investor in OpenAI, along with advanced microchip makers Nvidia and Arm. Stargate principals say they'll make an initial $100 billion investment, currently underway with 10 data centers under construction near Abilene, Texas, and are "evaluating potential sites across the country for more campuses as we finalize definitive agreements" and plans to invest up to an additional $400 billion over the course of Trump's second term. Trump said the effort will create 100,000 new U.S. jobs on the way to building "the physical and virtual infrastructure to power the next generation of AI."

OpenAI Chief Global Affairs Officer Chris Lehane characterized emerging artificial intelligence tools as representing "a quantum advancement, on the scale of the advent of electricity, that will impact how we live, how we work, how we engage with one another, even how we play." Lehane stated that "(Artificial intelligence) is inherently a productive technology" and called it "a tool to help people solve incredibly hard problems, whether it's in education, health care, sciences. These are ultimately tools to empower people to solve problems." Lehane also characterized the race to advance AI capabilities as a global competition, warning that "Up until relatively recently, there was a real sense that the U.S. had a material lead on the (Chinese Communist Party)" but that "it became clear last week when news emerged about (Chinese AI platform) DeepSeek ... that this is a very real competition and the stakes could not be bigger." He emphasized that "Whoever ends up prevailing in this competition is going to really shape what the world looks like going forward, whether we have democratic AI that is free and open or an authoritarian AI that is autocratic."

The infrastructure investments come with massive power requirements. During the press conference announcing Stargate, Ellison noted the Abilene data centers will be around 500,000 square feet per unit when complete, each capable of housing thousands of servers. Industry data reflects current pricing of about $7 million to $12 million per megawatt of IT load, or about $7 billion to $12 billion for a 1GW facility. A 2024 Goldman Sachs report details that on average, a ChatGPT query needs nearly 10 times as much electricity to process as a Google search. Goldman analysts said that at the time the report was issued, AI computational needs were consuming 1% to 2% of global power output but are on pace to increase by 160% by 2030.