This review article discusses the application of generative artificial intelligence (AI) to the de novo design of proteins. AI models, including large language models (LLMs) and diffusion processes, are used to generate novel protein sequences and structures with desired properties. The article categorizes these models as sequence-based or structure-based, highlighting recent advancements and comparing their effectiveness. Challenges remain in accurately predicting protein behavior and incorporating complex biological factors, but the field shows significant promise. The authors conclude by emphasizing the increasing accessibility of these tools and their potential for addressing various scientific and societal problems.