As artificial intelligence increasingly mediates human communication in the United States, understanding how individuals interpret, evaluate, and respond to algorithmic systems has become a defining competency of the 21st century. This paper introduces AI literacy as an expanded form of digital literacy that encompasses awareness of automation, data bias, and ethical reasoning in communicative contexts. Using mixed-methods research including survey data from 12,847 American adults and discourse analysis of professional and educational settings across 23 states, this study investigates how people develop critical understanding of AI-driven media environments. The findings reveal growing gaps between technical proficiency and ethical comprehension, with 73% of Americans using AI tools daily but only 34% demonstrating adequate understanding of algorithmic bias. The research emphasizes the need for educational frameworks that integrate media, data, and algorithmic literacy. The paper proposes a multidimensional AI Communication Literacy Model that redefines digital competence for an era of predictive algorithms and generative content, providing evidence-based recommendations for educational reform and policy development in the United States.