OpenAI brings local AI processing to UAE

OpenAI has introduced Inference Residency in the UAE, enabling organizations to run AI model inference on graphics processing units within the country. This service, which is available to eligible customers using OpenAI’s API, ChatGPT Enterprise and ChatGPT Edu, complements the company’s existing UAE Data Residency offering. Inference Residency ensures that GPU execution involving customer content can remain within the UAE, addressing concerns over where computational work is performed.

While data residency determines where customer information is stored, inference residency focuses on where the processing occurs. UAE customers must have data residency enabled in the same region to qualify. The UAE joins the US, Europe (EEA) and Switzerland as locations where OpenAI currently supports ChatGPT inference residency.

Eligibility is not automatic and requires engagement with OpenAI’s sales operation. Currently, GPT-5.2 is the model available for UAE inference residency, though GPT-5.6 is not supported. Image generation, internal search, ChatGPT Work and improved memory are not supported in the UAE configuration, though standard saved memories remain available.

Advanced Voice and external web search may also be unavailable or controlled by workspace settings. Notably, inference residency does not encompass all aspects of OpenAI’s processing; authentication, routing, analytics and external integrations may still occur outside the UAE. The UAE’s first international Stargate AI computing cluster of one gigawatt capacity is planned for Abu Dhabi, with 200 megawatts set for operation in 2026.