Kevin Dallas is responsible for all aspects of the EDB business globally. As CEO, Dallas accelerated the software company’s growth culminating in its $3.5 billion acquisition by Aptiv in 2022. In this exclusive interview, he delves into their recent research concerning sovereign AI platforms and the Middle East region.
How do you define “sovereign AI,” and why do you believe it will be the decisive differentiator for enterprises in the next five years?
True sovereignty means retaining full control over where your AI and data live, how they are accessed, and how AI is built and deployed. In practice, it means these assets are secure, compliant, and agile—available anywhere, anytime—across on-prem, multiple clouds, and the edge. It’s what allows enterprises to deploy AI wherever it creates value, without being locked into a single provider or constrained by black-box infrastructure.
And this isn’t just our view—it’s what the market is telling us. In our research with 2,050 senior executives across 13 global economies, 95% said becoming their own data and AI platform is a mission-critical priority in the next 100 days. This is not a regional trend or an industry fad; it’s a global business imperative. Sovereign AI is set to become the default operating model for enterprises worldwide because it enables them to move faster, stay compliant, and innovate without compromise.
Your research shows UAE and KSA outpacing global peers. What structural or policy decisions in these markets are enabling such rapid adoption?
On a national level, Saudi’s $77B HUMAIN initiative, the UAE’s Stargate investment, and other major public programs underscore both countries’ commitment to AI. Together they’ve created a capital-infrastructure-governance flywheel that compresses timelines for skills and platforms, accelerates enterprise readiness, and de-risks AI at scale.
In the private sector, UAE and KSA enterprise leaders are among the most Deeply Committed to data and AI sovereignty. In EDB’s 2025 research, leaders in these two regions scored 135 on the commitment index, well ahead of peers in the UK (75), the U.S. (111), and Singapore (124). This depth of conviction is turning the Gulf into a “sovereign hub,” where like-minded enterprises collaborate, exchange ideas, and accelerate their journey to AI and data sovereignty, together.
While these conditions have created fertile ground for AI adoption, our research highlights a universal truth: advantage isn’t inherent to any industry or region. It hinges on one variable—how deliberate and committed you are to data and AI sovereignty. Treat it as mission-critical and there’s a 90% likelihood you’ll be among the 13% of deliberate winners that pull ahead and sustain a measurable performance lead over the other 87%.
The report cites up to 5x ROI for deeply committed organisations. Can you share specific business functions or use cases where this ROI is most visible?
The clearest ROI shows up in data-intensive, regulation sensitive domains—customer-service copilots/bots, fraud and risk operations, autonomous logistics/supply chain, and telecom OSS/BSS.
But the real unlock isn’t the use case, it’s the approach. Our research shows that the top performers don’t just experiment; they deploy twice as many AI systems at scale and embed them 2.5x deeper into core workflows. They solve multiple pain points with a single intelligent application and run their agents “where the work is”—often in hybrid environments.
That’s why we talk about ROI as a portfolio of outcomes rather than a single metric. Innovation and efficiency together represent almost half the total return, but leaders also see gains in customer experience, faster time-to-market, cost reduction, and revenue growth. Across 15,000 simulations, we found a 0.93 correlation between sovereignty and economic/innovation/efficiency outcomes—meaning organizations that make sovereignty mission-critical are about 70% more likely to achieve compounding ROI.
What practical steps should enterprises—especially in sectors like banking, energy, and logistics—take to move from pilot AI projects to fully sovereign AI platforms?
To escape pilot mode and scale AI projects into mainstream production, enterprises should:
- Build for sovereignty, speed, and compliance. This means marrying AI and data together on a single platform and treating sovereignty as a non-negotiable from the start.
- Get data out of silos and embrace a hybrid approach. Run AI “where the work is” and move workloads to the most performant, cost-effective environment at any time.
- Make observability and governance a design principle. Ensure you have a single pane of glass to all your data wherever it lives, to ensure real-time insight into model performance, data movement, and compliance status across the entire data estate.
- Standardise on open technology. Anchor in open source platforms like Postgres to stay agile and extensible and avoid lock-in.
How do open-source technologies and cloud choices factor into building truly sovereign AI platforms?
Truly sovereign AI platforms are built on unified, hybrid-by-design infrastructures. This approach gives enterprises the agility and control needed to harness AI effectively and securely, without being hindered by data silos, exposing proprietary data, or locking into specific cloud providers.
Isolated single-cloud setups can’t meet true sovereignty requirements; real control means unified data access and complete observability across environments. Among the top cohort in our research, ~49% of workloads are designed as a cloud-like experience across flexible deployment options, enabling end-to-end functionality entirely within an enterprise’s self-hosted environment.
Across the global enterprise leaders we surveyed, open source is the one thing they standardize on: 81% of leaders say their future data infrastructure is open-source, and about one in three globally choose Postgres for new agentic/GenAI apps; across segments, up to 38% are selecting Postgres specifically for sovereign AI/data workloads.
Future Outlook and Global Implications – With the Middle East setting the pace, how do you expect global AI policies and enterprise strategies to evolve in response?
While both the public and private sectors in the Middle East have set a great example of pursuing AI sovereignty, geography isn’t the deciding factor, mindset is. In our research, geopolitics only accounts for 7% of the rationale for sovereignty adoption. The real drivers are agility (getting data out of silos) and compliance/security needs (together >75%).
Yes, we see higher percentages in places like the UAE and KSA (close to 30% of enterprises are already sovereign leaders), but we also see Deeply Committed enterprises across the U.S., Scandinavia, and Singapore. These companies view sovereignty as a competitive advantage, not just a regulatory checkbox. The U.K. is an interesting paradox and a good lesson for all: it has invested heavily in AI and infrastructure, but hasn’t produced comparable ROI. The spark was there, but execution and incentives have lagged.
The bigger point to remember for global enterprises is that: This is available to everyone. The 13% that have achieved sovereignty aren’t clustered in one region or industry — they simply made sovereignty a priority. The remaining 87% could unlock the same benefits if they committed to building sovereign, open, hybrid infrastructure now.
What is EDB’s long-term roadmap for enabling enterprises in this region to stay ahead in sovereign AI—both in terms of technology development and ecosystem building?
On the technology front, EDB is evolving Postgres into an AI-ready database, embedding vector search, fine-tuned retrieval, and real-time inference capabilities directly into the core. This ensures enterprises can build and scale AI workloads on sovereign, open-source infrastructure. Our focus is to give customers full control and sovereignty over where data and AI models are trained, stored, and deployed — whether on-prem, in local sovereign clouds, or in regulated hybrid environments.
This put Middle East enterprises front and centre—not only as adopters of AI, but as creators of sovereign, future-proof AI platforms rooted in open source, data sovereignty, and ecosystem self-sufficiency. Our continued investment and go-to-market in the region is delivered through collaborations with global partners such as NVIDIA, Red Hat, and IBM, as well as local ecosystem leaders including Saudi Business Machines, Gulf Business Machines, PowerM, and Ejada. These critical partnerships enable us to deliver capabilities that drive the C-Suite agenda and strategic imperatives for enterprises across the Gulf market.
EDB also offers world-class partner support that drives platform education and adoption —for example, through the EDB Postgres AI certification programs that are actively being rolled out across the Gulf region.
Image Credit: EDB
