The recent advancements in the field of watermarking and pseudorandom coding have significantly enhanced the robustness and applicability of these technologies. Notably, there has been a shift towards developing adaptive robust pseudorandom codes, which are crucial for watermarking applications in generative AI models. These codes now offer stronger guarantees against adaptive error channels, addressing a critical gap in previous constructions. Additionally, the concept of ideal pseudorandom codes has been introduced, which unifies pseudorandomness and robustness under a single definition, facilitating the construction of codes with linear information rates. In the realm of watermarking, there is a growing emphasis on localized and invisible watermarking techniques, particularly for AI-generated content. These methods not only improve imperceptibility and robustness but also enable the embedding of larger payloads, such as UUIDs with error correction, ensuring high accuracy even under severe image manipulations. The integration of deep learning models for localized watermarking has also shown promising results, offering new capabilities like the ability to locate and extract distinct messages from small, watermarked regions in spliced images. Overall, the field is progressing towards more sophisticated and versatile watermarking solutions that are better equipped to handle the complexities of modern digital content.