Abstract
The rise of low-cost artificial intelligence (AI) technologies offers significant potential for businesses globally, yet AI adoption remains uneven. What shapes this unequal adoption? While prior work attributes adoption patterns to demand-side factors including physical costs and complementary assets, we theorize that AI entrepreneurs' strategic choice to target specific markets creates both search and perceived-fit frictions for firms outside of those markets. We test this supply-side theory using unique data tracking the adoption of standardized AI tools—which require lower physical costs and fewer complementary assets—across approximately 88,000 firms worldwide from 2012 to 2023. Consistent with our targeting theory, firms based in non-English-speaking (i.e., non-targeted) markets are 16 percent less likely to adopt AI tools. These firms are even less likely to adopt AI when they serve different markets than the AI entrepreneurs—even though these firms observe a stronger relationship between AI adoption and subsequent performance. The findings highlight the strategic market choices of technology entrepreneurs as a key mechanism driving the unequal diffusion of frontier technologies.