Abstract
Artificial intelligence (AI) policy is increasingly used to align digital transformation with the Sustainable Development Goals by fostering decent work, innovation, and reduced inequality. Youth entrepreneurship is a key channel through which these objectives materialize, yet the micro level pathways linking AI policy to entrepreneurial entry among young adults remain under specified. Using five waves of China Family Panel Studies microdata from 2014 to 2022 and the staggered introduction of AI pilot cities, I estimate a two way fixed effects difference in differences model with rich individual, household, and regional controls. AI pilot status increases the probability of youth entrepreneurship, with baseline effects statistically significant at conventional levels and robust to event study tests of parallel trends, placebo reallocations, alternative time trend controls, and trimming of extreme values. Mechanism analyses show that AI policy operates by reducing relational spending and relaxing credit constraints, as verified by structural equation modeling and multiple mediation tests. Effects are stronger among healthier respondents, those with internet access, and rural residents. This study contributes to the international literature by providing micro-evidence from an emerging economy on how place-based AI policies can function as environmental equalizers. The findings suggest that the observed surge in youth entrepreneurship is structurally motivated by the substitution of digital rules for informal Guanxi and the replacement of physical collateral with digital credit scoring, offering scalable lessons for designing inclusive innovation policies.