Across the UAE and Saudi Arabia, 84 percent of organisations have deployed AI in at least one business function. Only 11 percent say it has made a meaningful difference to earnings. At Snowflake's Data for Breakfast event, enterprise leaders gathered to work out why, and what to do next.

[For more news, click here]

There is a peculiar kind of frustration building inside some of the Middle East's largest enterprises right now. It is not frustration born of inaction. Most large organisations across the GCC have been moving quickly on artificial intelligence, investing in tools, standing up pilots, and announcing transformation programmes with genuine intent. The frustration is something more specific: they have adopted AI widely, and yet the results, at the level of earnings, competitive advantage, and operational change that actually shows up in a business, have not followed at the pace they expected.

This is the gap that Snowflake's Data for Breakfast event, held recently for data and AI leaders from across the UAE and KSA, was designed to address. And the more you dig into the numbers, the clearer it becomes that the obstacle is not the AI itself. It is what sits beneath it.

The headline statistic from the GCC is striking: 84 percent of organisations are using AI in at least one business function. The follow-through is less impressive. Only 11 percent report a meaningful impact on earnings. That is not a technology failure. It is an architecture failure, and specifically a data architecture failure. AI systems, however sophisticated, produce reliable outputs only when they can draw on data that is well-governed, consistently structured, and accessible across the organisation. In most enterprises, that infrastructure simply has not caught up with the ambition of the AI layer sitting on top of it.

Moving Beyond the Pilot Trap

The phrase you hear most often in enterprise AI conversations right now is "moving beyond pilots." It sounds straightforward. In practice it describes one of the harder organisational transitions a large business can attempt: taking something that works in a controlled environment and embedding it so deeply into core operations that it starts to change how decisions get made and how revenue actually flows.

For Michel Nader, General Manager for META at Snowflake, the challenge is both commercial and strategic.

"Across the Middle East, AI has become central to national competitiveness, digital transformation, and future-ready economic growth. Organisations in the UAE and KSA are now focused on building data foundations that can support scalable, governed, and measurable AI. Data for Breakfast demonstrates how the Snowflake AI Data Cloud gives enterprises a practical path to unlock business value, strengthen decision-making, and advance secure agentic intelligence across the organisation."

The event brought together senior data and technology leaders from organisations across the region precisely to work through that practical path. Not the theory of enterprise AI, but the specifics: what the right data foundation actually looks like, how governance works in practice across complex organisations, and where the returns are most likely to materialise first.

The Insurance Company That Built Its AI Brain in Nine Months

One of the more compelling case studies at Data for Breakfast came from Emirates Insurance Company, one of the UAE's pioneering insurers, which has been running a significant digital transformation programme over the past year. The specifics matter here, because they illustrate what is possible when data architecture is treated as the foundation rather than an afterthought.

In nine months, Emirates Insurance Company built and went live with a new data platform on Snowflake, with the first go-live achieved in just three months. The platform has since improved quote-to-bind processes, created a more standardised operations workspace with stronger service level agreements, enhanced claims transparency, and supported faster decision-making across the business.

For Carlos Piedade, Head of Digital Strategy and Transformation at Emirates Insurance Company, the project was about building something that would outlast its initial use cases.

"With Snowflake, our goal was to build strong data foundations while delivering value to the business. Our data platform is becoming the brain of Emirates Insurance Company, helping us understand customers, brokers, performance, and risk, while Snowflake's capabilities and roadmap support our ambition to scale AI architecture, advance underwriting, and disrupt the UAE insurance sector."

That framing, the data platform as the "brain" of the organisation, is worth pausing on. It is the right metaphor for what enterprise AI actually requires. An AI system that cannot access consistent, trusted, well-governed data across an organisation is not a brain. It is a set of isolated reflexes. The transformation Emirates Insurance Company has undertaken is less a technology story than an organisational one: the decision to build the kind of data foundation that makes genuine intelligence possible.

Agentic AI and the Shift From Tool to Colleague

Much of the conversation at Data for Breakfast centred on what Snowflake calls agentic AI, and specifically on Snowflake Intelligence, its personal work agent for business users. The distinction from conventional AI tools is important and worth spelling out.

Most enterprise AI deployments to date have been what the industry describes as copilots: tools that assist a human user with a specific, bounded task. They are useful, but they are still fundamentally reactive. They wait to be asked. Agentic AI operates differently. It is designed to act across enterprise data and systems with a degree of autonomy, understanding the full context of an organisation's operations, and executing multi-step workflows through natural language.

Recent enhancements to Snowflake Intelligence make this more concrete. Users can now automate routine workflows through natural language instructions, connect their AI experience to tools including email, calendars, collaboration platforms, and CRM systems, and access the same capabilities on the move through an iOS mobile app. The governance and security architecture is built in at the foundation level, which matters in regulated industries and in the GCC's evolving AI compliance environment.

The vision, and it is one that a growing number of enterprise technology leaders in the region are starting to articulate, is of AI that does not sit alongside the work of a business but runs through it. That requires a data layer that is both comprehensive and trusted. It requires governance that is not bolted on but designed from the start. And it requires the kind of organisational will to invest in foundations rather than just features.

Investing in the Region's Digital Talent

Beyond the enterprise conversations, Snowflake used the event to reinforce its longer-term commitment to the region through its One Million Minds programme. Originally launched to build data and AI skills at scale, the programme is now expanding into Saudi Arabia with a specific aim: enabling thousands of Saudi learners by 2030 through applied training in data and AI, delivered with key national partners.

The initiative maps directly onto Saudi Vision 2030's priorities around workforce readiness, digital government, and economic competitiveness. In a region where AI ambition is running ahead of the talent base needed to execute on it, programmes that build practical capability rather than theoretical awareness are where the long-term value tends to compound.

The broader message from Data for Breakfast, if you take it as a whole, is that the GCC's enterprise AI story is entering a new chapter. The adoption phase, while far from complete, is no longer the central challenge. The central challenge is now depth: embedding AI into core operations with the kind of data architecture, governance, and organisational commitment that converts deployment into durable commercial advantage.

For the organisations that get that transition right, the 11 percent figure starts to look less like a ceiling and more like a baseline.

Related Articles:

Saudi Arabia Has Built the Right AI Foundations, According to PwC. Turning Them Into Profit Is the Next Test

Why Trust Has Become the Most Valuable Currency in MENA's Digital Payment Revolution

The Gulf Has Plenty of AI Ambition. Solutions+ and Inception Want to Make It Actually Work