Generative AI (GenAI) has rapidly moved from proof-of-concept to boardroom priority across the Middle East. While initial enthusiasm was anchored in the impressive capabilities of large language models (LLMs), enterprises are quickly learning that true competitive advantage demands more than just deploying the latest AI models.

The real value emerges when GenAI systems can access, understand, and act upon an organisation’s unique, constantly evolving data — and that is where Retrieval-Augmented Generation (RAG), and now agentic RAG, becomes mission-critical.

Why Standard LLMs Fall Short in the Enterprise

LLMs are trained on vast but static datasets, limiting their knowledge to information available at the time of training. In regulated or fast-moving sectors, these limitations manifest as outdated or incomplete responses, raising both compliance risks and operational frustrations.

For Middle Eastern enterprises facing evolving regulations and dynamic market conditions, these gaps are not just technical shortcomings — they can directly impact revenue and reputation.

RAG: Transforming LLMs into knowledge powerhouses

RAG bridges this critical gap by connecting AI models directly to diverse, real-time enterprise data. This not only keeps outputs current and reliable, but also boosts performance on tasks that require domain-specific knowledge or regional context.

Additionally, RAG frameworks eliminate the need for continuous and expensive re-training of core models, streamlining scalability and reducing time-to-value as business use cases and data sources evolve.

The market appetite reflects these benefits: global RAG spend is projected to soar from $1.2bn in 2024 to over $67bn in 2034, expanding at a compound annual growth rate of nearly 50 per cent.

Agentic RAG: The next evolution — from answers to autonomous action

With agentic RAG, enterprises are empowered to move beyond information retrieval to intelligent action. Rather than responding passively to user prompts with static outputs, agentic AI systems can autonomously plan, negotiate, execute, and optimise tasks — all while grounded in the latest organizational knowledge.

Imagine AI assistants dynamically managing supply chain schedules, automatically resolving customer queries, or orchestrating employee onboarding — all with minimal human intervention but maximum compliance, consistency, and strategic alignment. For CXOs, this represents not just a step-change in productivity but a true leap in enterprise agility and innovation capacity.

Agentic systems also introduce new dimensions of responsibility. Their autonomy and proactivity require strong frameworks for governance, transparency, and trust — especially as Middle Eastern governments advance national AI strategies and data protection laws.

A strategic playbook for Middle East CXOs

Embracing agentic RAG is as much a leadership mandate as it is a technology upgrade. Consider the following strategic actions:

• Build a unified data and infrastructure foundation
  Advance past silos by investing in robust, cloud-native data architectures. Standardised governance and privacy-first practices ensure that GenAI systems remain compliant with local regulations (such as GDPR and PDPLs) and are equipped for regional growth.
• Prioritise AI governance from day one
  Autonomous systems raise new questions of accountability. Establish ethical guidelines, audit trails, human oversight, and scenario-testing as non-negotiables. Transparency and responsible AI are essential to align with both stakeholder expectations and regulatory mandates.
• Develop true workforce-AI synergy
  The Middle East is witnessing exponential growth in demand for AI and ML skills. To fully harness agentic RAG, invest in upskilling programmes and nurture talent capable of translating business needs into AI outcomes. Encourage a culture of collaboration between human and machine.
• Start with impactful pilots
  Adopt a “test-and-learn” mindset. Initiate agentic RAG pilots in high-value domains — such as automated customer support, dynamic supply chain adjustments, or internal policy management. Use clear KPIs and ROI metrics to guide rapid iteration and scaling.
• Tie GenAI to tangible business outcomes
  Anchor every AI initiative in measurable value. Whether it’s reducing decision latency, enhancing the customer journey, or driving cost efficiencies, agentic RAG works best when it’s solving real business problems for real people.

Call to action: Lead the evolution, don’t watch from the sidelines

For Middle East CXOs, the status quo is no longer enough. Leading organisations are already transitioning from generic LLM deployments to bespoke, agentic RAG-powered ecosystems where intelligence is grounded, decisions are automated, and opportunities scale with data.

The challenge is not just to keep pace, but to set the pace. Elevate GenAI discussions from IT operations to foundational business strategy. With agentic RAG, you’re not just enabling smarter automation — you’re building an enterprise that is resilient, adaptable, and primed for the future of work.

Now is the time for Middle East business leaders to champion this evolution: secure your data, empower your teams, govern your AI, and reap the rewards of truly intelligent, action-oriented enterprise systems.

The writer is the MD Gulf at Dell Technologies.

When Charles Haresnape took the helm at Gatehouse Bank eight years ago, the focus was on transforming it from a predominantly commercial player into a retail-driven Shariah-compliant bank. Today, the UK-domiciled, PRA and FCA-regulated institution serves clients from around the world, with the Gulf Cooperation Council (GCC) a key growth market.

“We’re a fully Shariah-compliant bank in the UK, but you don’t have to require Shariah compliance to be our customer. That’s the basis upon which we provide our services, and we’re proud of it,” Haresnape says. “Of all my customers who are based outside the UK, 20 per cent are in the GCC and that’s growing.”

Consistent GCC appetite for UK property

For GCC-based investors, the UK remains a preferred destination for property investment, regardless of macroeconomic uncertainty.

“We’re seeing consistent demand for UK property. In fact, the demand has held up strongly, even more so than perhaps our domestic demand,” Haresnape notes. “When people look back over many recessions and economic cycles, they see UK property as faring very well. It’s very consistent and recovers quickly.”

He adds that falling interest rates are further supporting investor sentiment: “People see it as a good opportunity to get in at a lower rate than previously.”

Gatehouse specialises in residential investment rather than commercial, with an average home finance size of £300,000 to £400,000 — enabling investors to spread risk across multiple properties.

Buy-to-let and build-to-rent both in focus

Haresnape says GCC clients are active in both individual landlord and institutional segments — an unusual combination for a bank of Gatehouse’s size.

“We’re very rare in the UK to offer both buy-to-let and build-to-rent,” he explains. “For buy-to-let, which is more the individual landlord or small company, we’re seeing increasingly people building larger portfolios to give more flexibility. The trend in the UK, which also applies to the GCC, is clients wanting several properties rather than just one or two.”

The build-to-rent segment caters to large-scale overseas investors — often sovereign wealth funds or family offices — pooling hundreds of millions into UK residential property. “We currently manage over 11,000 properties in the UK on behalf of those funds,” he says. Past investors have included the Kuwait Investment Authority, alongside global private equity players such as the Carlyle Group and TPG.

Read: Navigating the new tax environment for GCC family offices

Strengthening the UK–GCC financial corridor

Haresnape sees strong regulatory alignment between the UK and GCC, making cross-border finance smoother. “In Saudi Arabia, for example, they’ve been fast followers of UK regulation standards, and there’s much more consistency now. In Dubai and Kuwait, regulation has also increased, which is good for customers,” he says.

This consistency builds trust, he adds, with customers across the GCC better understanding the protections and processes around financial services.

ESG and Islamic finance: a natural alignment

Islamic finance’s restrictions on certain industries and its risk-sharing principles naturally align with environmental, social and governance (ESG) values. Gatehouse is leveraging this to attract both ethical retail savers and impact-driven institutional investors.

“We were founder members of the United Nations Principles for Responsible Banking, and we fundamentally believe in our ESG principles ; they’re not just window dressing,” Haresnape says. “People increasingly want their finance providers not to be involved in certain types of activity — arms trade, gambling, alcohol, drugs — and we make a complete statement that we don’t invest in those areas.”

The bank has been carbon neutral for five years, plants trees for every new savings account opened, and focuses on energy-efficient building standards in the properties it finances. “We’re testing solar panels, increasing the use of ground-source heat pumps, and prioritising environmentally friendly building methods,” he says.

This ESG proposition resonates strongly with younger savers. “In our 2024 survey, 83 per cent of respondents aged up to 24 said ESG was an important factor in deciding on financial products,” Haresnape notes.

Regional investment hotspots

While some GCC investors buy London property for personal use, Haresnape says most prefer higher-yield regional markets such as Manchester, Liverpool and Birmingham.

“House prices there are lower than in London, and rental levels remain healthy, so yields are stronger. Even institutional investors take the same approach,” he explains.

Growing the GCC footprint

Looking ahead, Gatehouse is planning deeper engagement in the GCC. “We’ll be there more frequently, have more people on the ground, and increase broker relationships in the region,” Haresnape says. “The GCC is driving the growth of Islamic finance quite heavily, and the demand is increasing significantly.”

With majority ownership by Kuwaiti institutions, including the Kuwait Investment Authority as its largest shareholder, Gatehouse Bank is well-positioned to capitalise on this growth.

“The future of Islamic finance in the GCC is strong and so is the appetite for UK property investment from the region,” Haresnape concludes.

UAE-based tea brand FiLLi Cafe has partnered with COFE Cloud to digitise and optimise its operations as part of an expansion plan covering the UK and Asia, the companies said on Thursday.

The agreement centres on a customised mobile app built on COFE Cloud’s digital solutions platform, which is aimed at helping coffee and beverage brands scale efficiently while improving customer experience.

The brand, which operates in the UAE, US, UK, Oman, Qatar, Mauritius, Nepal, India and Pakistan, is looking to accelerate its international growth with COFE Cloud’s data-driven tools, operational streamlining and customer engagement capabilities.

“We are proud to support FiLLi, one of the UAE’s most iconic café brands, on their next chapter of growth,” said Ali Al Ebrahim, founder and CEO of COFE. “By bringing our advanced digital infrastructure to their expansion strategy, we aim to redefine what’s possible for regional brands scaling globally.”

FiLLi to get a fully customised app

COFE Cloud’s system will give the retail brand a fully customised app to reflect local preferences across different markets and improve operational efficiency, revenue growth and brand loyalty.

“Partnering with COFE Cloud marks a significant milestone in our digital journey,” said Rafih FiLLi, founder and CEO of FiLLi Cafe. “This collaboration enables us to seamlessly scale our operations while staying true to our brand promise—delivering world-famous Zafran Chai and Global Street Food with excellence, speed, and personalisation.”

Founded in 2004 as a single café in the UAE, FiLLi has grown into a global brand known for its modern take on street food fusion and signature chai.

COFE, founded in 2018, operates in Kuwait, Saudi Arabia, UAE and Egypt, with further expansion planned.

Dubai Electricity and Water Authority (DEWA) reported a record revenue of Dhs14.6bn for H1 2025, marking a 6.9 per cent increase year-on-year.

The company also announced a dividend payment of Dhs3.1bn, payable in October 2025.

The half-year financial results highlight strong growth across key metrics.

DEWA posted a profit after tax of Dhs2.9bn, up 13.2 per cent compared to the same period in 2024.

EBITDA reached Dhs7bn, reflecting a 5.3 per cent rise, while operating profit stood at Dhs3.7bn, increasing 12.6 per cent.

Operating cash flow surged 61.3 per cent to Dhs9.2bn.

DEWA’s Q2 highlights

In Q2 alone, DEWA recorded a profit after tax of Dhs2.4bn, up 25.8 per cent year-on-year, with revenues of Dhs8.6bn, an increase of 9.8 per cent.

EBITDA and operating profit grew 11.9 per cent and 24.8 per cent respectively.

DEWA currently serves 1,292,487 customer accounts, a 4.81 per cent increase over the same period last year.

Saeed Mohammed Al Tayer, MD and CEO of DEWA, said, “We are proud to report DEWA’s strongest-ever financial results for both the second quarter and first half of 2025 … Our record operating cash flow of Dhs9.2bn, up 61.3 per cent year-on-year, underscores the resilience of our business model.”

Al Tayer credited the leadership and vision of Dubai’s rulers and noted continued investments in state-of-the-art infrastructure totaling over Dhs230bn to date.

Looking ahead, DEWA expects stronger revenue and profit contributions in H1 2025, driven by seasonal demand and continued investments in clean energy, digital infrastructure, and water desalination in line with Dubai’s Green Economy vision

Governments worldwide face mounting fiscal pressure: rising debt, volatile revenues, and growing public expectations.

Amid this complexity, they are expected to act faster, spend smarter, and enhance trust in public institutions. Artificial intelligence (AI) presents a once-in-a-generation opportunity to redefine how public resources are planned, allocated, and accounted for. Yet its potential in public finance remains largely untapped.

Governments are beginning to integrate AI into fiscal operations – from optimising budgets and improving forecasts to automating audits and fraud detection. These early efforts hint at a bigger prize: the strategic use of AI to redesign the fiscal policy cycle itself.

The question is no longer whether AI can help, but how fast and how well governments can scale its use responsibly.

AI’s role in modern public finance

AI is transforming how governments manage public finances. It enhances decision-making in fiscal policy and resource allocation, strengthens risk management, streamlines operations, and improves citizen-facing services.

By moving beyond simple automation, AI enables real-time data analysis, dynamic resource targeting, and proactive risk identification.

These capabilities are already being applied across public finance to support:

• Macroeconomic and fiscal forecasting: AI is transforming traditional econometric methods by using machine learning (ML) and deep learning (DL) to process vast, unstructured datasets. This improves forecasting accuracy and enables real-time “nowcasting”. For instance, the Australian Taxation Office uses ML models to forecast tax revenues, while South Korea’s Ministry of Economy and Finance produces daily updates on the national treasury balance using AI. In the UAE, the Ministry of Finance is enhancing revenue forecasting and compliance through AI, while initiatives like Smart Dubai embed intelligent tools in services such as digital payments and smart procurement.
• Budget planning and expenditure monitoring: AI modernises budgeting processes by automating data handling and applying advanced analytics. ML enhances the accuracy of expenditure baselines, supports policy cost estimation, and enables evidence-based fiscal decision-making. For example, the Australian Department of Veterans’ Affairs uses predictive models to simulate lifetime fiscal impacts of beneficiaries and assess policy options, while France’s DGFiP applies ML to identify municipalities at financial risk, evolving from historical data analysis to predictive forecasting.
• Public spending reviews: AI is strengthening spending review processes by analyzing large and complex datasets to identify trends, evaluate programme effectiveness, and inform resource reallocation. ML and DL extend beyond traditional analytics to uncover deeper insights and automate recommendations. For instance, the UK Treasury employs HMT-GPT to assess budget proposals and support long-term funding reviews, while Canada’s Department of Finance uses AI to evaluate the impact of public spending and guide reallocation decisions.
• Accounting, control, and fraud detection: AI-driven automation and anomaly detection are making internal financial controls more efficient. Tools using NLP, ML, and DL, can rapidly process documents, identify irregularities, and strengthen oversight. Denmark employs AI to monitor subsidy disbursements and flag anomalies, and the UK applies ML to detect fraudulent benefit claims with improved speed and accuracy.
• Citizen engagement and service delivery: AI is redefining how public finance institutions interact with citizens. Chatbots and language models enhance accessibility, automate responses, and improve transparency. The U.S. Internal Revenue Service uses AI-powered voice and chatbots to reduce inquiry wait times, while Ireland’s Department of Finance uses AI to draft tax manuals and summarize legal documents, making government communication more accessible.

These examples underscore AI’s growing role across the fiscal value chain, but also reveal a gap: AI is informing decisions, not making them.

Why prescriptive AI remains elusive

Prescriptive AI – the ability to recommend or make decisions – remains rare in public finance.

The reasons are complex: lack of explainability, unclear accountability, and unresolved ethical concerns. Should an AI system decide how public funds are distributed or which programs face cuts? What if its recommendations reflect bias or flawed assumptions? Who is accountable when things go wrong? These are not just technical questions – they are governance questions. Addressing them is key to unlocking AI’s next frontier in fiscal policymaking.

What’s holding AI back?

Despite its promise, AI adoption in public finance faces five persistent barriers:

• Lack of strategic alignment with institutional priorities: Many institutions lack a top-down, structured approach to identifying AI use cases that directly support national priorities or institutional mandates. This leads to fragmented, opportunistic, or siloed implementations and limits the ability to demonstrate strategic value, especially when impact tracking is focused solely on cost or operational efficiency.
• Outdated infrastructure and fragmented data ecosystems: Legacy IT systems and non-integrated data sources hinder effective AI model development. High-quality, interoperable data is essential but often inaccessible or trapped in bureaucratic systems resistant to integration. These challenges are particularly acute in regions where coordination across agencies remains limited. In the GCC, efforts to unify public finance platforms – often led by sovereign wealth funds or centralised finance ministries – highlight the growing need for shared standards and interoperable systems.
• Capacity and culture gaps: AI deployment requires more than technical expertise. It demands a culture that embraces innovation and adaptive decision-making. Many institutions lack digital capabilities, face internal resistance to change, or operate within risk-averse environments where experimentation is discouraged. Regional actors such as the Arab Monetary Fund have highlighted the need for stronger institutional coordination and innovation ecosystems to advance digital finance transformation across the Arab region.
• Ethics, security, and transparency concerns: As AI begins to shape sensitive fiscal decisions, such as allocating benefits or reallocating funds, issues of fairness, legality, and accountability become critical. Without clear rules, AI can produce biased outcomes or breach financial regulations. Weak cybersecurity may expose sensitive fiscal data, threatening national security and eroding trust. Public finance professionals and citizens must understand how AI insights are generated and used. Opaque algorithms or poorly communicated logic risk undermining both legitimacy and public confidence.
• Absence of robust evaluation and ROI frameworks: AI returns are harder to measure and often intangible in the short term. This makes it challenging to prioritise and scale promising pilots. Without clear methodologies to assess impact – including efficiency gains, accuracy improvements, and equity outcomes – AI programmes struggle to secure sustained funding and political backing.

A strategic path forward

To move from pilots to purpose-driven AI adoption, governments should focus on five priorities:

• Anchor AI in core fiscal strategy: Define AI priorities top-down, aligned with institutional and national fiscal and development goals. Focus on areas where AI advances mandates such as revenue mobilisation, spending efficiency, or compliance.
• Invest in infrastructure and people: Build modern cloud infrastructure, hire skilled data engineers, and provide ongoing training to unlock AI’s full value beyond isolated pilots.
• Strengthen data governance: Establish strong data governance frameworks to improve data accessibility, quality, and interoperability, while safeguarding privacy and promoting ethical use.
• Measure what matters: Track cost-benefit metrics alongside accuracy, compliance, equity, and public confidence to capture AI’s true impact on fiscal management.
• Embed safeguards: Require model transparency, independent audits, and clear accountability frameworks before AI tools influence high-stakes fiscal decisions.

Read: The AI imperative: 5 steps to transforming public sector services

The time to act is now

AI is not just a technological upgrade – it is a fundamental shift in managing public finance. Governments that embed it strategically will unlock unprecedented agility, precision, and transparency. Moving beyond advisory roles, prescriptive models can drive smarter, faster, and more accountable policy decisions.

But this power demands caution: without rigorous transparency, fairness, and accountability safeguards, such systems risk bias and unintended consequences that could undermine trust. Done responsibly, AI can help governments anticipate shocks, improve policy outcomes, and enhance public confidence.

For MENA countries, where fiscal reform and economic diversification are top priorities, the stakes are even higher. With the right investments, governance, and institutional commitment, the region can not only catch up, but lead in shaping the future of public finance.

Naman Sharma and Pedro Marques are partners, and Rayane Dandache is a manager at Kearney Middle East & Africa – Financial Services Practice.