This paper asks whether adopting artificial intelligence (AI) lowers firms’ required returns—or instead raises perceived uncertainty—and why effects differ across firms and markets. Building on finance theories of information risk and control, we develop a framework in which AI affects the cost of capital through three channels: risk estimation (forecastability of cash flows), risk transformation (tail exposure and correlated failure risk), and governance credibility (legibility, auditability, and accountability of AI-driven decisions). The central prediction is conditionality: AI can improve operating performance yet generate little average decline in the cost of capital when opacity or control concerns offset cash-flow gains. Financing benefits arise when disclosure and governance make AI use verifiable and constrain downside risk, with particularly strong implications for debt markets and mature firms. The framework motivates a governance-interacted staggered-adoption difference-in-differences design and clarifies why prior evidence on technology adoption and financing outcomes is mixed. It also yields sharp implications for FinTech valuation, where AI is often the core production technology and capital pricing hinges on credible oversight.