Abstract
As corporate finance architectures rapidly evolve, the integration of Generative Artificial Intelligence (GenAI) and autonomous intelligent agents is fundamentally restructuring financial planning and analysis (FP&A), enterprise risk modeling, and capital allocation strategies. This paper presents a multi-industry empirical investigation evaluating the operational and financial performance outcomes of automated AI agents deployed within modern corporate finance units. Drawing upon Dynamic Capabilities Theory and the Technology-Organization-Environment (TOE) framework, our findings indicate that real-time predictive analytics powered by multi-agent LLM architectures yield an 84.6% reduction in forecast error rates and accelerate quarterly financial closing cycles from weeks to hours. However, the adoption of autonomous financial algorithms introduces novel systemic risks, including algorithmic bias, hallucination risks in cash flow forecasting, and heightened cybersecurity exposure. This study establishes a quantitative governance framework to balance technological efficiency gains with robust risk mitigation protocols.