G42 has received approval from the United States to import advanced AI semiconductors, marking a major shift from planning to execution across the UAE–US AI collaboration. The decision signals strengthened bilateral trust and enables the rapid scaling of secure, high-performance AI infrastructure across both nations.
The approval directly accelerates flagship initiatives such as Stargate UAE, the 1-gigawatt AI compute cluster G42 is building for OpenAI alongside Oracle, Cisco, NVIDIA, and SoftBank Group. This cluster forms part of the wider 5-gigawatt UAE–US AI Campus, designed to deliver massive compute capacity and low-latency inferencing for regional and global workloads.
The move also expands G42’s collaborations with US hyperscalers and chipmakers, including Microsoft, AMD, Qualcomm, and Cerebras.
All advanced chip deployments will operate under the Regulated Technology Environment (RTE) — a compliance framework built by G42 and aligned with US Department of Commerce and BIS standards to ensure secure global diffusion of American technology.
G42 Group CEO Peng Xiao said the approval “sets a new benchmark for secure, high-performance compute,” adding that G42 will maintain infrastructure symmetry between the UAE and the US to reinforce trust. Khaldoon Khalifa Al Mubarak said the decision reflects the depth of the UAE–US relationship and the shared vision for technology as a driver of stability and long-term cooperation.
G42 now operates three of the world’s Top500 supercomputers, including the #2 and #3 systems in the region, and recently launched its Maximus-01 system in New York — ranked 20th globally. Its AI infrastructure footprint spans Abu Dhabi, France, and multiple US states including California, Minnesota, Texas, and New York.
