Emirates NBD is committing roughly 50% of its technology budget to artificial intelligence, the clearest signal yet that UAE lenders treat AI as core infrastructure rather than an experiment. The Dubai-based, majority government-owned bank disclosed the figure at last week's Dubai FinTech Summit 2026, alongside more than 150 AI initiatives, of which 65 are already in production.

How far along is the deployment?

In June, Emirates NBD ranked first in the Evident AI Index for Banks – Middle East & Africa, a survey of more than 100 companies, including insurers and payment processors, run by London-based market intelligence firm Evident. The index assesses how banks across the regions deploy AI and adapt operations to the technology. Evident noted that Emirates NBD "explicitly" links AI to enterprise productivity, operational modernisation and the UAE's national AI agenda.

Rankings measure intent as much as results, and a first-place score does not by itself prove returns. The bank's more concrete move came last month, when it became the first bank in the MENA region to deploy OpenAI's GPT models on local, in-country infrastructure — running large language models (AI systems trained to generate and analyse text at scale) while meeting UAE data residency, privacy and regulatory compliance requirements.

Emirates NBD plans to integrate OpenAI's models across operations to improve customer experience, strengthen risk management and financial crime prevention through advanced analytics, automate processes and raise employee productivity.

What does the bank's leadership claim?

Miguel Rio-Tinto, Emirates NBD's Group Chief Digital and Information Officer, set out an aggressive timeline: "Banking will transform more in the next five years than it transformed in the last 30 years."

He added: "Generative AI and advanced frontier models are so powerful that they are opening up new possibilities for automation in banking, enabling capabilities that were previously out of reach across customer service, internal processes and employee support. I believe AI will be the key determinant in shaping the competitive landscape of the future."

Every transformation promise carries a cost, and Rio-Tinto's own numbers show where it lands first: cybersecurity. As AI models grow better at identifying software vulnerabilities, critical security flaws requiring evaluation at the bank have surged from roughly 1,000 per month in early 2025 to about 15,000 last month. The bank is now using a frontier OpenAI model defensively, to test and identify weaknesses in its own systems before attackers do.

Why is the threat rising now?

The regional context has sharpened. Since the outbreak of the US-Iran war on February 28, pressure on operational resilience across the Gulf has intensified. In the UAE, daily attempted cyberattacks reportedly surged from around 200,000 to as many as 700,000 amid escalating geopolitical tensions, according to a Help AG State of the Market report.

In July, the UAE thwarted a series of sophisticated cyberattacks targeting its financial sector, which sought to compromise digital systems and technical infrastructure through phishing campaigns and malicious software. The UAE's Cyber Security Council warned that cybercriminals are increasingly using AI to devise more sophisticated attack methods. Neither the council nor the banks have published full damage assessments, which makes independent verification difficult.

What does this change for the wider sector?

Other lenders are moving in parallel. Commercial Bank of Dubai (CBD) unveiled a series of strategic technology partnerships at the summit:

• Microsoft, to support wider AI adoption
• Silent Eight, for AI-powered compliance and risk management
• Comarch AI Wealth, to develop its digital wealth platform and investment offerings
• Zamp, for intelligent automation
• Harvey, Sirion and Flank, to improve legal and governance infrastructure

Thomas Cherian, CBD's chief technology and services officer, said: "We are moving away from traditional AI applications towards adopting agentic AI, which is creating opportunities to rethink and redesign existing business processes." Agentic AI, unlike conventional generative AI, can make decisions and execute tasks autonomously with minimal human intervention.

Cherian estimates AI agents could handle 55% to 60% of operational tasks, with human oversight retained to approve outcomes. CBD also aims to retain ownership of the agents it develops and the intellectual property behind them — a caveat that suggests banks want the upside without ceding control to vendors.

Is the state pushing this agenda?

The government is doing precisely that. In April, the UAE launched a national initiative to move 50% of all federal government operations, services and tasks to agentic AI systems by 2028. The following month, Dubai announced a two-year action plan to accelerate agentic AI adoption across the emirate's private sector, aiming to secure a "competitive edge" in the future global economy.

Ambitious targets and heavy bank investment point to genuine transformation. Whether five years will outdo three decades, as Rio-Tinto claims, depends on execution that no index ranking can yet guarantee.