Foundation models, such as large language models (LLMs), diffusion models, and vision-language models (VLMs), have emerged as powerful tools for various applications. In this talk, we will explore the exciting potential of foundation models in revolutionizing wireless communications and networking. On AI for communications (AI4COM), we will focus on three key aspects, namely "learning to compress", "learning to estimate", and "learning to optimize". Meanwhile, we will also discuss effective methods to enable foundation model empowered AI services at the network edge.