AI across the Gulf
Industry deep dive: Banking and financial services

By Nizar Hneini, Saumitra Sehgal and Luca Turba

Banking and financial services report among the most mature sectors in AI adoption across the GCC. A majority (62%) report having fully documented digital and AI strategies, with alignment to national visions and use of KPIs tied to AI-related business impact both higher than the cross-industry average. Investment momentum remains strong: 86% of banking respondents expect AI budgets to increase, including a substantial share anticipating significant increases.

On paper, the financial sector looks ready to scale. In reality, delivery remains uneven. 34% report that AI is already scaling across the enterprise, while 66% believe there’s more to be done. The gap between strategic intent and operational execution is the defining challenge for the sector today.

Where AI is creating value today

Banking organizations expect AI to deliver value across efficiency, decision-making and customer experience. Faster and better decisions and productivity gains are among the most frequently cited outcomes. Enhancing customer experience is also reported as a driver for prioritizing AI adoption across departments, in line with cross-industry findings.

Across the sector, banks and financial institutions appear to be prioritizing internal process transformation first: governance, reporting and compliance automation, before extending AI into customer facing applications. This sequencing reflects both the regulatory sensitivity of the sector and the complexity of integrating AI meaningfully into customer journeys. Half of respondents report tracking AI performance using business-linked KPIs, and initiatives aligned with clear governance and performance measurement practices are more likely to continue beyond pilot stages.

End-user adoption: where scaling succeeds or stalls

Our survey finds that trust in AI outputs among banking leaders is generally high, but more polarized than in other sectors. While most report full or partial trust in AI-generated outputs, 12% report no trust, compared with 9% across industries. This cautious profile is consistent with the sector’s regulatory environment, where consequences of AI errors are more visible and consequential than in most other industries.

The consequences of AI errors carry real regulatory and reputational risk, which makes decision makers appropriately demanding about reliability before they commit to scale. Adoption challenges therefore relate to integrating AI into existing workflows rather than to lack of awareness or willingness to use AI tools.

What is holding AI back from enterprise scale

Data quality is the most frequently cited constraint on scaling AI in banking, with 58% of respondents reporting data quality issues as a major barrier. Fragmented data environments and legacy core systems limit reuse and expansion of AI use cases.

Beyond data quality, the findings point to fragmented ownership of AI initiatives as an underlying obstacle. Our survey reports that 32% of banking respondents describe business and technical team collaboration as mostly siloed — the highest of any sector and above the 23% cross-industry average. Without clear cross-functional accountability, use cases that succeed in one department will rarely translate into enterprise-wide deployment.

Technology integration challenges are also widely reported (39%), alongside weak cross-functional collaboration (38%). Seamless integration into customer journeys — at both the experience design and technology architecture level — remains a distinct challenge, with 34% of banking respondents citing the ability to integrate AI into existing workflows as a missing organizational capability. Security and data protection concerns add further complexity, with 26% citing cybersecurity as a major obstacle. Together, these constraints explain why so many institutions find themselves caught between successful pilots and consistent enterprise-scale delivery.

Banks and financial institutions across the GCC have been among the earliest adopters of AI, but early adoption has not automatically translated into enterprise scale. The institutions that piloted AI tools now face a harder problem: how to move from isolated use cases to an operating model in which AI is embedded across the business.

That transition requires a different kind of commitment. It means reimagining processes and operating models end-to-end, rather than layering AI onto existing structures. Three priorities stand out for leaders navigating this next phase.

1. Review internal processes and reimagine operating models
Early AI wins have largely come from automating discrete tasks within existing structures. The next level requires stepping back to ask which processes should be redesigned, not just optimized.

2. Resolve fragmented ownership
AI scaling stalls when accountability for business outcomes is fragmented. Organizations need cross functional deployment teams with explicit ownership of business outcomes, not just technology delivery. Without this, even well-funded initiatives lose momentum as they move from pilot to scale.

3. Make output quality a governance priority
In banking, the stakes of AI errors are high. Hallucinations, false positives, or compliance monitoring carry direct regulatory and reputational consequences. Active monitoring of AI outputs is not optional; it’s a prerequisite for earning the organizational trust that sustained scaling requires.

To explore the full data and insights behind these trends, download the complete AI across the Gulf: From ambition to scalable impact report here