Over 10,000 of the 34,500 large companies worldwide (usually 30%) have already committed to becoming their own data and AI platforms. And within just three years, over 95% of global companies say they want to do the same, from Japan and India to the US, EMEA and MEA, regardless of where they operate. This is not just a local trend. It's a global mandate. And it's accelerating.
The first three waves of digital transformation took over a decade to rebuild the industry, but the agent revolution could unfold within three years. That's not just speculation. This is the views of 2,050 executive leaders in 13 economies surveyed in the 2025 Global Survey. Their consensus is clear. The time to act is now.
This is a story of two futures. The advantage is profound as 13% of companies already have it right. They report five times the ROIs of their fellow ROIs and deploy agent and generative AI features on a double scale. Perhaps, more specifically, these leaders are 250% confident in their ability to remain at the forefront of the industry for the next three years. Their choice of platforms produces measurable economic outcomes, not merely strategic.
Meanwhile, the other 87% aren't necessarily struggling, but they're behind. Their returns on AI investments are, at best, only a small portion of what leaders are generating. At worst, they are five times less productive and still rely on legacy infrastructure, unable to meet the rising expectations of their customers, regulators and shareholders.
Convictions of moving to sovereign data and AI infrastructure are shared around the world. Large companies in Scandinavia, the US, Saudi Arabia, the UAE and Japan are rapidly building their own platforms. And of the 13% of high performers, the most concentrated success stories have emerged in Germany, Saudi Arabia, the United Arab Emirates and the United States. These regions show what is working and why.
Rule 1: Mission-Critical Sovereignty enables scalable agent AI
What distinguishes 13% is not just what they are doing, but how they think. Almost 90% of these organizations share one critical mindset. They view sovereignty over AI and data as mission-critical. They believe that all unstructured, unstructured data is accessible in real time, regardless of where it is accumulated. Prioritizes eliminating silos, enforcing compliance, and creating an environment where data and AI can work seamlessly together.
This belief is not theoretical and has been tested. Our study spanned over 15,000 simulations combining 15 agents and Genai features with seven economic indicators and 35 data infrastructure variables, consistently outperformed by organizations focusing on real-time sovereign access and control. Statistically, this pattern of success shows a very high predictive score (Chaid and Regression: 0.982), confirming that this idea is simply not useful. That's essential.
These leaders do not treat AI and data as separate areas. They merged them into a single sovereign engine that encourages operations in more than a dozen business areas. They built a platform that doesn't bother you where your data resides, so that your design ensures access, security, compliance and scale. As a result, they move faster, innovate more, outperform their competitors.
Every day, an average of 58 new businesses begin this transition. That's the scale of change we're witnessing. Questions for those who haven't moved yet are easy. Do you have a thriving infrastructure and thinking?
Rule 2: Compliance, Observability, Security and Agility should be designed with
Sovereignty is not just about ownership. It's about architecture. The most successful organizations don't just talk about sovereignty. They designed it into four interdependencies: compliance, observation, security and agility.
They have created an environment where AI can interact with the real world without compromising data integrity or violating regulatory obligations. For example, financial forecasting agents should pull from CRM, customer experience, sales funnels, ledger data, and external benchmarks. Edge agents may need to ingest signals from vehicles, sensors, or telecom infrastructure. Sovereign AI needs infrastructure to support all of this in a unified and compliant way.
Of those deeply committed, over 40% operate on a truly hybrid infrastructure. It supports full observability across the environment, from the cloud to on-prem, and provides centralized control through one pane of glass. This allows you to simultaneously respond to competitive threats, regulatory changes and customer demands. They reduce their dependence on global infrastructure providers, build architecture flexibility, and recognize benefits that are more than twice as much as their colleagues' innovation and efficiency.
These are not abstract advantages. These are measurable business outcomes that deteriorate over time, and are what many call sovereign flywheels. With unified control, innovation is faster. With compliance built in, agility increases. Once observability is complete, optimization becomes constant. And when security is unique, trust is an asset.
This is why the sovereign movement is growing so rapidly. It's not just the right thing. That's beneficial.
Redefine the success of the Agent and Gen AI era
What's happening is not just AI, but a global transition. Sovereign AI. More than ever, companies are realizing that “having” AI is not enough. True success comes from owning infrastructure, data, and intelligence end-to-end.
Open source platforms like PostgreSQL are powering this shift. Flexible, hybrid, combat-tested Postgres now allows businesses to build resilient systems that unify data across AI and the environment, accelerate and scale sovereign strategies.
Because at the end of the day, it's not sovereignty of data or AI. Building a right architecture with the right mindset and the right principles is the definition of success in a world where AI is no longer an option.
The race continues. The winners have appeared. And the rules have become clear.
For more information about EDB, please see here.
