This paper provides the first micro-level evidence from a nationally representative U.S. household survey on individuals' reliance on artificial intelligence for financial advice. Using weighted binomial logit models, we examine how digital readiness, risk preferences, investment sophistication, and behavioral frictions shape AI advice adoption. Mobile banking usage and risk tolerance significantly increase reliance on AI advice, while ownership of securities and cryptocurrencies reduces it. Behavioral factors-financial anxiety, time burden, and perceived financial constraint-emerge as the strongest and most robust predictors. The results indicate that AI financial advice primarily functions as a behavioral and cognitive support technology rather than a substitute for financial expertise, with important implications for financial inclusion and consumer protection.