Large language models like GPT-3 are increasingly becoming part of human communication. Through writing suggestions, grammatical assistance, and machine translation, the models enable people to communicate more efficiently. Yet, we have a limited understanding of how integrating them into communication will change culture and society. For example, a language model that preferably generates a particular view may influence people's opinions when integrated into widely used applications. This dissertation empirically demonstrates that embedding large language models into human communication poses systemic societal risks. In a series of experiments, I show that humans cannot detect language produced by GPT-3, that using large language models in communication may undermine interpersonal trust, and that interactions with opinionated language models change users' attitudes. I introduce the concept of AI-Mediated Communication–where AI technologies modify, augment, or generate what people say–to theorize how the use of large language models in communication presents a paradigm shift from previous forms of computer-mediated communication. I conclude by discussing how my findings highlight the need to manage the risks of AI technologies like large language models in ways that are more systematic, democratic, and empirically grounded.