OpenAI has launched its “Inference Residency” service in the United Arab Emirates. With the new implementation, eligible customers’ AI model inference operations can be carried out on GPUs located within the country’s borders. The UAE has thus become the third region, after the United States and Europe, to offer “inference residency” for ChatGPT.

Model Inference with OpenAI Will Take Place Within the UAE

Unlike the existing “Data Residency” implementation, Inference Residency determines not only where data is stored, but also where the GPU operations through which the AI model processes customer content are carried out. To benefit from the service, data residency must also be enabled in the same region. While eligible ChatGPT Enterprise and Edu customers can use this option, regional processing support is also available for certain API services in the UAE.

E-Commerce and Retail Companies Will Also Be Able to Benefit

OpenAI’s new infrastructure also allows e-commerce and retail companies that meet the eligibility requirements to run supported AI workloads in the UAE. Customer content covered by the service includes prompts, files, conversations, and embeddings data generated from this content. The local GPU processing option is available not only to sectors with specific data-location requirements, such as finance, healthcare, education, and government, but also to e-commerce and retail businesses using artificial intelligence.

GPT-5.2 and Feature Limitations on ChatGPT

According to OpenAI’s official information, GPT-5.2 is currently available in ChatGPT workspaces configured with inference residency in the UAE. Image generation, built-in search, ChatGPT Work, and enhanced memory are not supported, while standard saved memories can continue to be used. In addition, some non-GPU processes, such as authentication, routing, and analytics, may be carried out outside the region.

AI Adoption Is Accelerating in the UAE

Farouk El Hamzawi, Head of Enterprise for OpenAI MENA, stated that organizations in the UAE are rapidly adopting artificial intelligence and that the government’s long-term technology vision is supporting this process. El Hamzawi said that with the Inference Residency service, eligible customers are given greater control over where their AI workloads are run, while stronger regional assurances are provided to organizations scaling their use of artificial intelligence.