OpenAI has introduced inference residency in the UAE, giving eligible ChatGPT Enterprise and Edu customers the option to run AI model inference on GPUs located within the country.
The move makes the UAE the first market in the Middle East and North Africa (MENA) to offer both data residency and inference residency for ChatGPT.
The development could strengthen the UAE’s position as a regional hub for enterprise AI. It also gives organisations in regulated sectors greater control over where their AI workloads are processed.
What is inference residency?
Data residency and inference residency address two different parts of AI processing.
Data residency controls where customer content is stored when it is not being used. Inference residency controls where the AI model processes that content on GPUs.
In simple terms, data residency answers “Where is my data stored?”
Inference residency answers “Where does the AI process my data?”
OpenAI says inference residency keeps GPU-based model inference on in-scope customer content within the selected region. However, some CPU-based processes, such as authentication, routing and analytics, may still take place outside the UAE.
UAE joins US and Europe
The UAE is now one of only three regions where OpenAI offers ChatGPT inference residency.
The other two are the United States and Europe, covering the European Economic Area and Switzerland. OpenAI’s current documentation lists the UAE as a supported inference region alongside these markets.
The service is available to eligible ChatGPT Enterprise and Edu customers. However, organisations must first enable UAE data residency before they can use inference residency.
OpenAI introduced UAE data residency in November 2025. At the time, the company said eligible organisations could store customer content in the UAE to meet local data protection and regulatory requirements.
Why the move matters for UAE businesses
The new capability could be particularly important for sectors where data location is a key concern.
These include:
- Financial services
- Healthcare
- Government
- Education
- Other highly regulated industries
For these organisations, keeping AI model execution inside the country can offer an additional layer of control.
Moreover, the move may make it easier for businesses to adopt generative AI while meeting internal data policies and local regulatory requirements.
OpenAI describes inference residency as an additional location guarantee for GPU execution. It does not, however, mean that every part of the AI service operates exclusively inside the UAE.
GPT-5.2 is currently the main limitation
There are some restrictions for UAE-resident workspaces.
At present, GPT-5.2 is the only model supported for ChatGPT inference residency in the UAE. OpenAI also says that image generation and internal search are not available under the UAE configuration. External web search can remain available, subject to workspace controls.
In addition, third-party services remain outside OpenAI’s residency guarantee.
For example, data sent to external apps, MCP servers, web-browsing providers or other third-party tools is handled under those providers’ own terms.
Therefore, inference residency should not be viewed as a guarantee that every piece of processing stays within the UAE.
OpenAI’s UAE enterprise footprint is growing
The latest move follows a broader expansion of OpenAI’s presence in the UAE.
When the company expanded data residency globally in November 2025, it listed the UAE among the countries where eligible business customers could store customer content locally. OpenAI said it had more than one million business customers globally at the time.
The UAE has also emerged as an important market for OpenAI’s regional expansion.
The company has previously announced partnerships and deployments involving major UAE organisations and institutions. These include Abu Dhabi-based AI company G42, sovereign wealth fund Mubadala, property developer Aldar and fintech company Tabby.
Universities such as Khalifa University, Mohamed bin Zayed University of Artificial Intelligence, NYU Abu Dhabi and UAE University have also been linked to OpenAI’s UAE ecosystem.
What OpenAI says about inference residency
OpenAI’s guidance draws a clear line between data storage and model execution.
“Inference residency adds an additional location guarantee for GPU execution on customer content.”
The company also makes clear that the feature does not cover every type of processing.
Authentication, authorisation, routing and other CPU-based operations may still occur outside the selected region.
A step towards deeper AI sovereignty
The announcement comes as the UAE continues to invest heavily in AI infrastructure and digital sovereignty.
In May 2025, OpenAI announced Stargate UAE, a planned 1GW AI infrastructure project in Abu Dhabi. The project formed part of the broader Stargate initiative.
The latest development adds another layer to that expansion.
Data residency gives organisations a local place to store their AI content. Inference residency gives them a local place to run GPU-based AI processing on that content.
That distinction could become increasingly important as governments and businesses place tighter controls on where sensitive AI workloads are processed.
For the UAE, the move also strengthens its case as a regional base for enterprise AI. OpenAI says it plans to expand inference residency to more regions over time.
What this means for employers and HR teams
The impact also extends beyond IT departments.
HR teams increasingly use AI for recruitment, employee support, learning and workforce analytics. In regulated organisations, however, concerns around employee data and information security can slow adoption.
Local AI processing could help organisations address some of those concerns.
As a result, UAE employers may have more room to explore enterprise AI tools while keeping key customer content and GPU-based processing within the country, subject to the limits of the residency framework.
