Saudi Arabia's AI ambitions shift from pilots to national infrastructure

Saudi Arabia's AI ambitions are entering a new phase. Across government entities, critical industries and private enterprises, AI is moving beyond pilot projects and proofs of concept into the systems that shape daily operations and long-term national competitiveness.

As thousands of technology leaders gather for LEAP 2026, the conversation is increasingly focused on practical questions of implementation, infrastructure, talent and governance.

Underneath all of the decisions that AI enables sits physical infrastructure, and data centres have their own limits. Sami Alfaraj, MEA Head of Technology at Submer, explains those buildings have outgrown their old category. "When 66 of Vision 2030's 96 strategic objectives depend on data and AI, the infrastructure running those systems is no longer an IT asset. It is a national infrastructure, and it deserves to be designed like it," he said.

For decades, facilities were designed first and filled with servers later. "AI inverts this. The accelerator roadmap now dictates the building," Alfaraj said. One rack of today's hardware draws over 100 kilowatts, more than ten older racks combined, and throws off heat that air cooling cannot shift.

Rahul Misra, Senior Vice President and Managing Director for MEA at IFS, works with industries where short delays cost real money. "For the Gulf states, where asset-heavy industries dominate the economy, industrial AI embeds intelligence directly into operational environments, transforming how organisations run their most critical processes. The biggest difference is autonomous operational orchestration," he said. "Oil and gas, manufacturing, utilities, and logistics operate at scales where coordination delays directly impact margins and reliability."

Systems that manage suppliers, stocking and maintenance scheduling across connected assets stop small problems from multiplying through a supply network and cut the hours lost due to unplanned downtime.

The same tools have sped up how quickly attackers get in, and Maher Jadallah, Vice President for Middle East and Africa at Tenable, says spotting a flaw is the easy half.

"Discovering a vulnerability is like identifying a symptom, and exposure management is the cure," he said. The scoring system commonly used highlights up to 60 percent of flaws as high or critical, but only about 3 percent of flaws pose the risk of being exploited by attackers. Attacks are often conducted based on a single flaw and attackers combine minor flaws, open cloud configurations, excessive permissions, then attack something valuable.

"By the time an alert triggers, an automated attacker may have already explored the network," he said, arguing that the answer is pre-emptive fireproofing.

With AI seeping into everything technology has to offer, and when the conversation moves to the value of AI, Lakshimi Narayanan, General Manager at Mannai Information Technology Saudi Arabia, shared "The organisations getting real value decided what would change before what to buy."

Those companies establish performance metrics before implementing, rebuild the process around the technology rather than retrofitting an old process with new technology,and hand it over to the department responsible for the results.

"A decision made multiple times a month is far more valuable than a strategically fascinating one made twice a year," he said.