Emirates NBD will become the first bank in the Middle East and North Africa to deploy OpenAI's GPT models on in-country inference infrastructure in the UAE, the bank announced on October 19, 2026. Inference is the step where a trained model answers a request, so in-country inference means prompts and answers are processed inside the UAE.
The bank ties the choice to regulation. It says in-country inference supports local expectations on data residency, privacy and operational resilience, and presents the deal as aligned with the UAE's National Strategy for Artificial Intelligence 2031.
Planned uses include customer engagement across digital channels, risk and compliance analytics, automation of knowledge-heavy workflows and employee productivity. Emirates NBD also plans to connect the models to its internal systems, APIs and enterprise data to build new digital services and strengthen risk monitoring and financial crime controls.
Several details are missing. The release names no data centre operator, no model versions, no timeline and no budget. It also does not say whether the enterprise data used by these connections stays in the country along with the inference. Those answers decide how far sovereign goes in practice.
Why this matters for banks and fund issuers
Where AI runs is becoming an architecture decision, not only a contract clause. Banks now choose between global cloud APIs, in-country cloud inference like this deal, and models on their own servers, and each option trades capability against control and running cost. The institution has to be able to explain that choice to its supervisor (see how LabForty sets up local AI around your boundaries). The first evidence will be which Emirates NBD services go live on the new setup.
