Investing in AI: Generative AI Adoption by Country

43% of U.S. workers now use generative AI at work. France sits at 28%. Italy, 26%. 📊

That’s not noise. That’s two economies absorbing the same technology at fundamentally different speeds.

The adoption numbers alone would be worth watching. But the productivity data makes it sharper. U.S. workers report saving about 6.1% of their weekly hours through AI, roughly 2.5 hours. Germany and France come in around 4.8%. Spread that difference across millions of knowledge workers, week after week, and what starts as a firm-level edge becomes a country-level productivity gap.

Three things are driving the compounding:
→ U.S. firms measure and reward output gains faster
→ Procurement cycles are shorter, so new tools get adopted quicker
→ Middle managers actively push AI into day-to-day workflows, rather than waiting for top-down mandates

A recent NBER working paper picks up early (non-causal) signals that these time savings are already showing up in real productivity growth. Worth noting: this data was captured before practical agents hit mainstream workflows. The current gap is a floor, not a ceiling.

Fair counter: EU privacy regulation may build genuine trust advantages in certain verticals. But trust is a feature. Adoption velocity is the compounding engine.

For vertical AI companies, this plays out in a specific way. U.S.-based design partners give you faster feedback loops, faster iteration, faster product-market fit. That kind of advantage compounds quietly until one day it isn’t quiet at all.

Early signs suggest this gap is widening, not closing.