Everyone is claiming to be AI-native right now. Very few actually are.

The distinction matters — not as a label, but because it determines how a company scales, where its moat comes from, and whether it survives the next wave of model improvements.

The test

An AI-enabled company uses AI tools to do existing work faster. Remove the AI and the company still functions — it's just slower.

An AI-native company is built around AI as its primary operating capability. Remove the AI and the company stops working. The unit economics, the team structure, and the product architecture all assume AI at the core.

This isn't a technology question. It's a business model question.

Why it matters for investors

AI-enabled companies will face increasing margin pressure as AI capabilities become commodity. If your competitive advantage is "we use AI to do X faster", that advantage compresses as everyone gets access to the same models.

AI-native companies build moats differently:

Data moats — the company gets more valuable as it accumulates proprietary data that improves its models. Each operation makes the next one better.

Workflow moats — the company's AI is deeply embedded in customer workflows. Switching requires rebuilding, not just replacing a tool.

Knowledge moats — the company's AI encodes operational know-how that took years to develop. A new entrant can't replicate it by training on public data.

The failure mode

The hardest part of evaluating AI-native companies is that founders don't always know which one they are.

A founder building an AI-enabled product can pitch it as AI-native. The difference only becomes visible when you ask: what happens when your competitors have access to the same foundation models?

If the answer is "we have better data" or "we're embedded in the workflow" — that's a moat. If the answer is "we'll figure that out" — that's not.

This is the first question we ask. The answer shapes everything else.