OpenAI has expanded its residency options in the United Arab Emirates to include local model inference, allowing eligible customers to keep GPU execution of prompts and responses inside the country. The feature is now available for ChatGPT Enterprise, ChatGPT Edu and the OpenAI API, making the UAE one of only three locations worldwide alongside the United States and Europe where OpenAI supports full inference residency.
This builds directly on the data residency at rest that OpenAI previously enabled in the UAE. With inference residency enabled, in-scope customer content is processed on GPUs located in the region rather than being sent to infrastructure outside the country. The company positions the move as a way for enterprises, public-sector bodies and education providers to meet local compliance, governance and operational requirements while scaling AI use.
Farouk El Hamzawi, OpenAI’s Head of Enterprise for MENA, said the launch gives customers greater control over where AI workloads run and represents another step toward broader compute residency in the UAE. The feature aligns with the country’s National AI Strategy 2031 and rapid enterprise adoption of generative AI.
Inference residency requires data residency to be enabled in the same region. For the UAE, current limitations include support only for GPT-5.2 in ChatGPT and the API. Image generation, internal search, ChatGPT Work, improved memory and GPT-Live are not available for workspaces configured with UAE inference residency. Standard saved memories remain supported.
The expansion addresses a common enterprise concern: data-at-rest residency alone does not guarantee that prompts and model outputs stay within jurisdictional boundaries during processing. By constraining GPU inference locally, OpenAI reduces that gap for customers subject to UAE data-protection rules or internal policies.
The UAE has emerged as a high-priority market for OpenAI, with strong government support for AI and growing institutional use. Local inference residency strengthens the company’s competitive position against cloud providers that already offer in-region model hosting. Its longer-term value will depend on how quickly additional models and features are brought under the residency umbrella and how seamlessly customers can adopt the constraints without losing capability.
