Dell Technologies’ Mohammed Amin on scaling sovereign AI, from data residency to operational trust
Amin shares what comes next for sovereign AI, the implications of agentic systems and how organisations can close the gap between AI ambition and operational delivery.
09 September, 2026
TT
16
The Gulf’s AI ambitions are moving into a new phase. From the UAE’s push to build massive AI infrastructure to Saudi Arabia’s plans to rapidly expand data centre capacity, billions are being committed to the computing power needed to support the region’s digital economies. The challenge now is turning that capacity into secure, scalable and trusted AI.
For Mohammed Amin, SVP, CEEMETA at Dell Technologies, sovereignty increasingly extends beyond where data is stored. It is about who controls the data and models, how AI systems are governed and whether critical infrastructure can remain resilient as AI becomes embedded across government and enterprise operations. He speaks to Gulf Business about what comes next for sovereign AI, the implications of agentic systems and how organisations can close the gap between AI ambition and operational delivery.
Sovereign AI is entering a new phase across the region. How is the conversation changing?
Sovereign AI is scaling rapidly across Central and Eastern Europe, the Middle East, Türkiye and Africa (CEEMETA), fuelled by ambitious government mandates, national digital strategies and significant capital investment aimed at building greater local digital independence.
Today, the conversation is moving beyond where data sits to whether governments and enterprises can build AI environments that are secure, observable, resilient and ready for the next generation of AI.
That shift matters because sovereign AI is no longer only a policy or data residency discussion. It is becoming an infrastructure and delivery question: can organisations access advanced compute, retain control over data and models, govern how AI is used and scale innovation responsibly?
Across the region, organisations are navigating a unique landscape in which they must balance strict regulatory compliance with the agility needed to innovate.
What kind of investments are driving this growth across the region?
Across the region, sovereign infrastructure is scaling rapidly, with the UAE anchored by the multi-gigawatt Stargate AI campus and Saudi Arabia planning to develop more than 6GW of data centre capacity by 2034 to support sovereign AI workloads.
IDC expects AI spending in the Middle East, Türkiye and Africa (META) to reach $14.6bn by 2028, growing at a compound annual growth rate of 34 per cent.
This level of investment will only translate into long-term value if organisations can pair sovereign compute capacity with trusted data control, governance and resilience.
Why does sovereign AI need a new trust architecture right now?
With AI increasingly becoming the backbone of daily enterprise and government workflows, traditional IT security models are no longer sufficient to guarantee safety or compliance.
Sovereign AI needs a new trust architecture that provides the ability to prove where data lives, how models are trained or deployed, who or what can access the environment, how workloads are governed and how services remain resilient as AI becomes embedded in everyday operations.
This becomes even more urgent as AI moves towards agentic systems that can plan, coordinate and execute tasks across environments. These systems will operate continuously, interact with multiple applications and support decisions in more complex ways.
Sovereign AI infrastructure therefore needs to support not only data control, but continuous, observable and governed AI operations.
For CIOs and public-sector technology leaders looking to turn sovereign AI ambition into systems that can deliver, what are the key shifts that matter most?
First, start where the impact is highest. Organisations should begin where sovereign AI can deliver measurable value fastest, rather than treating it as an abstract infrastructure project. The focus should be on a clear outcome, such as faster services or better use of national data.
Second, match infrastructure to data sensitivity and workload demand. Not every workload runs in the same place. Sovereign AI environments must bring compute, storage, networking, resilience and energy considerations into a single operating model.
Third, build governance into that operating model. Transparency, accountability and alignment with regional regulatory frameworks from bodies such as the UAE’s AIATC and Saudi Arabia’s SDAIA need to extend directly into how systems are designed, deployed and used.
Fourth, prepare for agentic readiness by ensuring infrastructure can support always-on workloads across on-premises, edge and cloud environments.
Finally, make resilience part of sovereignty from the start. Organisations need the ongoing ability to restore trusted data, validate systems and maintain critical operations under pressure.
What ultimately defines success for organisations in the next phase of sovereign AI?
Sovereign AI is not just about where AI runs, but whether the full environment can be trusted to deliver.
The next phase of sovereign AI in META will be led by businesses that treat infrastructure, governance and resilience as one architecture, rather than three separate workstreams.
Ultimately, the winners will be those that can successfully move from data residency to operational trust, and from AI ambition to measurable, responsible delivery.





















