ABU DHABI, Oct. 26 (EconoTimes) — G42 Group Chief Executive Peng Xiao presented artificial intelligence services as a possible UAE export business, linking the ambition to a planned Abu Dhabi campus that would provide large-scale computing capacity.
Stargate UAE is a one-gigawatt computing cluster planned within a five-gigawatt UAE–U.S. AI campus in Abu Dhabi. G42 says it is building the cluster for operation by OpenAI and Oracle; Cisco, NVIDIA and SoftBank Group are also involved. G42 expects the first 200 megawatts to begin operating in 2026, but has not given a specific date that year.
G42 says the campus will cover 10 square miles and use nuclear, solar and natural gas to supply power. The company says the project is intended to support scientific discovery and innovation in health care, energy, finance and transportation. Its announcement gives no project cost or export revenue forecast and names no overseas customers.
The National separately reported that Xiao said the planned infrastructure could potentially reach about four billion people through the UAE’s connectivity. That is a potential reach estimate, not a count of current users or confirmed customers. The National provided no methodology for the figure.
The UAE government has set a separate public-sector goal. A framework announced in April 2026 aims to apply agentic AI models in 50% of government sectors, services and operations within two years. The government says the initiative is intended to lower operating costs and increase productivity. It does not establish an official target to replace energy exports with AI exports, so Xiao’s export ambition remains his stated vision.
The government says the UAE launched its AI Strategy 2031 and appointed a Minister of State for Artificial Intelligence in 2017. The announcements describe planned computing capacity and domestic AI adoption, but give no figures for commercial returns or actual overseas demand. The first 200-megawatt cluster, expected to begin operating in 2026, is the next stated project milestone.
