Algorithms and Capital: How AI Reshapes Corporate Finance Decisions

This paper develops a unified conceptual framework for how artificial intelligence (AI) reshapes corporate-finance decisions. We model AI as a bundle of capabilities (prediction, monitoring, disclosure, automation) operating through five channels: information/forecasting, agency/monitoring, disclosure/intermediation, operational automation, and model risk. These channels affect the core policy margins of capital structure, debt maturity, liquidity, investment, payout, risk management, issuance terms, and M&A quality. We introduce two AI-specific primitives, information precision and model-risk severity, into trade-off, pecking-order, agency, and disclosure-based theories. We derive propositions P1-P18 on how AI can drive either lean, efficient balance sheets or precautionary reversals with higher buffers and hedging. We then map these propositions into an empirical research agenda based on measurable constructs for AI adoption, exposure, governance, forecast quality, and model risk, and outline identification strategies using staggered adoption, regulatory shocks, and crosscountry data. The framework highlights dynamic, heterogeneous effects and the central role of AI governance in determining financing outcomes.