Realtime AI News
Emirates NBD to Run OpenAI GPT Models In-Country, a First for MENA Banks
Emirates NBD will run OpenAI's GPT models on servers inside the UAE, making it the first bank in the Middle East and North Africa to use in-country inference. The bank plans to apply the models to customer engagement, risk and compliance analytics, knowledge-heavy workflows and employee productivity, though no launch date or cost has been disclosed.

Emirates NBD will run OpenAI's GPT models on servers located in the United Arab Emirates, according to a PCMag Middle East report published on October 19. The arrangement makes the Dubai-based lender the first bank in the Middle East and North Africa to use in-country inference, meaning the step in which a model processes a request and produces an answer happens inside the country rather than in a data center abroad.
For a bank, keeping inference at home is less a technical preference than a precondition for data residency and regulatory compliance. The companies say the setup is designed to help Emirates NBD meet local rules on data residency, privacy and operational resilience, so prompts and business data are handled within the UAE.
The bank has identified four areas of use: customer engagement across digital channels, risk and compliance analytics, automation of knowledge-heavy workflows, and employee productivity. The models will also be connected to internal systems, APIs and enterprise data, which the bank expects to translate into faster development of digital services, better personalization, and tighter risk monitoring and financial crime controls. The release gives no timeline for any customer-facing feature.
The deployment is framed within the UAE's National Strategy for Artificial Intelligence 2031. Both companies say it aligns with the national AI agenda and reflects the growing availability of sovereign AI, meaning AI that runs under local governance and regulatory control. Banks are a natural test case, since they hold some of the most sensitive data and answer to strict supervisors.
Miguel Rio-Tinto, the bank's Group Chief Digital and Information Officer, described the plan as capability building rather than experimentation. "AI is not an experiment for us, it is a core capability that will define how we build, serve and scale," he said. Farouk El Hamzawi, Head of Enterprise at OpenAI MENA, cast the first-in-region approach as a benchmark for how financial institutions can deploy AI securely. OpenAI will supply the models, tools and the UAE-based inference.
The deal extends OpenAI's broader push into banking, including ChatGPT for Financial Services, and it lands as Google Cloud pitches Gemini Enterprise to the same sector. Core banking operations are becoming the next contested ground for foundation model vendors, and data residency terms are often the first gate such deals must pass.
Emirates NBD says 97% of its financial transactions and requests already take place outside branches, a digital-first customer base that gives the AI work room to scale. Yet the release does not say when the deployment goes live or how much it costs, leaving the actual production load and return on investment unclear for now.
Three questions are worth watching: when the system carries real production traffic, how its per-request inference cost compares with calls routed abroad, and whether regulators treat it as a reference template for other banks in the region. Clear answers on all three would turn sovereign AI in finance from a policy phrase into a replicable engineering blueprint.
Why it matters
If it works, the deployment turns sovereign AI from a policy slogan into a replicable template for regulated industries, and could push other regional banks to treat in-country inference as a baseline requirement.
Nearby Updates
All10/19, 15:53
Anthropic pitches Congress its own frontier-AI bill with a 50-death or $1B damage halt threshold
Anthropic has pitched Congress a frontier-AI bill of its own drafting that would allow models to be halted only after causing 50 deaths or $1 billion in damage, according to Dealroom. The proposal shifts the regulatory trigger from pre-deployment review to measurable harm, opening a debate over whether those thresholds are set far too high.
10/19, 15:00
Vercel: AI Gateway spending up 13-fold since December as customers rotate models
Vercel's latest AI Gateway Production Index shows spending on its model-routing service up 13-fold since December, while the average cost per token fell 47% over the same period. Open-weight models handled 71.5% of September tokens, and nearly one in two teams switched their most-used model during the month.
10/19, 12:47
DeepSeek ships V4.1-Flash-1017, claiming faster tasks and $0.2 coding runs
DeepSeek released V4.1-Flash-1017, an upgrade to last month's V4.1-Flash, claiming faster task completion, shorter answers and sharply lower costs. The company says a single coding task now costs about $0.2, less than a tenth of U.S. rivals, and cites a self-published DeepSWE v1.1 score of 77.0.
10/19, 11:00
OutSystems report finds only 26% of enterprises ready for AI agent conflicts
An OutSystems report produced with CIO Dive and KPMG finds that 82% of executives are confident their organizations can manage AI agents in production, but only 26% feel well prepared to resolve conflicts across agents, systems, and automated workflows. The survey of more than 850 IT decision-makers also shows sharp regional gaps in governance readiness, with Asia-Pacific ahead of Europe and North America.