What 500 leaders from AI-native companies told us about infrastructure

AI-native growth depends on something far less visible than the products these businesses sell: the infrastructure underneath them.

We surveyed 500 AI-native business leaders across the United States to understand how infrastructure choices affect their ability to scale. The findings show that infrastructure has moved from a background operational concern to a strategic decision that shapes how fast you ship, which customers you can reach, and whether you can expand into new markets. 

In this video, I explore three findings from the research and what they mean for teams planning their next phase of growth.

‍

1. Infrastructure is already limiting what AI-native teams ship

For most teams, this is already a problem today. 76% of respondents said infrastructure limited their ability to ship features in the past year. That's three in four companies whose roadmap was set by their infrastructure instead of their customers.

That pressure is building. 94% expect their compute needs to grow over the next six to 12 months, and 37% expect growth of 10x or more.

At that pace, capacity planning becomes part of product planning. The fastest-moving teams map out compute alongside their roadmap, rather than treating it as a follow-on step.

2. Fragmentation carries a measurable cost

The average AI-native business runs on 4 infrastructure providers and tools. Running multiple providers can be a deliberate choice, made for coverage, flexibility, or leverage in negotiations. But the cost is real, and it's often unmeasured.

60% of respondents said they lose more than a quarter of their engineering time to orchestration and infrastructure overhead. That's time spent integrating compute, data, and tooling across the stack instead of building models and products. 

The question for teams is whether their stack delivers the capabilities and commercial flexibility they need without adding seams their engineers have to manage.

3. Sovereignty and compliance are infrastructure decisions

Regulatory and sovereignty requirements are now a practical part of infrastructure design. Nearly all respondents (96%) are navigating at least one regulatory or sovereignty requirement, and 77% say data residency is a very or critically important consideration when choosing a provider.

Retrofitting sovereignty and compliance into a live stack can be slower, harder, and riskier than building them in from day one. In regulated markets, that readiness can determine which customers you can serve. 

Three takeaways for AI-native teams

The findings point to three priorities AI-native teams can act on now:

• Plan capacity for the scale you expect. Make compute part of roadmap planning from the start.
• Measure what fragmentation costs you. Know how much engineering time your stack takes to run, then decide where integration can reduce that overhead.
• Design for sovereignty from the start. Map regulatory and data residency requirements before you decide where workloads run.

‍

In my view, teams that make these decisions early will be better positioned to ship faster and sell into more markets as they scale. 

Explore the full findings, including five actions for leaders planning ahead, in the "State of AI Natives 2026" report.

‍