For the past three years, Wall Street and headlines have been fixated on a single infrastructure story: hyperscaler dominance. Anyone who bought into AWS, Azure, Google Cloud in 2023 was rewarded. But investors still crowding into that trade in Q1 2026 may be chasing it in the rearview mirror. A quieter rotation is taking shape inside enterprise infrastructure, one driven not by abandoning the cloud, but by a new enterprise priority: control.
A growing number of CFOs and CIOs are coming to the same realization: as AI workloads graduate from experimental pilots to always-on, production-scale systems, the question is no longer simply where compute is cheapest. It’s where AI can run securely, predictably, and under sovereign governance.
Reports last week of a potential 15–20% price increase on high-performance AI cloud instances rippled through the tech sector. For years, the prevailing wisdom was simple: move everything to the cloud. But AI is changing that calculus. Variable pricing models, jurisdictional uncertainty, and rising regulatory pressure are turning infrastructure into a strategic boardroom issue.
That shift is pulling capital toward what some investors are calling the “sovereign AI trade,”: the companies building platforms that allow enterprises to operate modern AI factories across hybrid environments — spanning cloud, on-premises, and sovereign deployments — with consistent governance and control.
“For enterprise leaders managing massive data portfolios and accelerating AI initiatives, the threat of cloud lock-in is no longer just a budgetary footnote—it’s an existential availability and compliance risk.” Quais Taraki, CTO of EnterpriseDB (EDB), a key player in this shift. “Beyond cost, this is about sovereignty. Organizations need to control where their data lives and how AI is deployed, without being locked into any single provider or environment.”
This sentiment is echoing through earnings calls this quarter. Companies are realizing that building modern, containerized AI factories on internal hardware, often powered by NVIDIA, is not only far more cost-effective than indefinite renting but also provides the control and agility they need to govern and scale their most strategic assets.
This is where private equity firms, like Bain Capital, are placing their bets. Not on the hyperscalers themselves, but on the companies supplying the tools behind the private infrastructure shift.
EDB has moved beyond being just a “database company” and now positions itself as a sovereign data and AI platform. The thinking is straightforward. If you are running complex workloads, especially in highly regulated verticals such as BFSI, healthcare, or the public sector, the challenge is no longer just scaling AI without exploding costs. It is ensuring that AI operates with full assurance: where data lives, who governs it, and how models are deployed.
“The market is starting to separate the companies that rent their intelligence from those that actually own it,” Taraki notes. “The winners in this 2026 rotation will be the ones who treat their AI stack like a capital asset, not just an operational expense. This shift from cloud-first to hybrid-by-design is as much about controlling the economics as it is about ensuring governance, compliance, and long-term competitive value.”
This is what investors are starting to price in as the “sovereign premium.” As data residency rules tighten across Europe and Asia, being able to say, “We know exactly where our data lives,” is turning into a real competitive advantage. That level of certainty is something the public cloud simply cannot promise. A sovereign, hybrid platform can.
For investors, that shift demands a closer look at the SG&A lines across tech-heavy portfolios. Are companies renting their intelligence, or do they actually own it?
The repatriation trade is real. And the shift toward hybrid infrastructure is only accelerating.
