The field of content authentication and watermarking is rapidly evolving, driven by the increasing need to verify the origin and authenticity of digital content. Recent research has focused on developing innovative watermarking techniques that can effectively detect and trace AI-generated content, including images, text, and audio. These techniques aim to address the challenges posed by the proliferation of generative AI models, which can produce highly realistic but potentially misleading content. Notable advancements include the development of robust font watermarking methods, multi-layer network frameworks for certifying human-originated content, and plug-and-play parameter-intrinsic watermarking for neural speech generation. These innovations have significant implications for applications such as copyright protection, content moderation, and digital forensics. Some noteworthy papers in this area include FontGuard, which introduces a novel font watermarking model that modifies fonts by altering hidden style features, and P2Mark, which proposes a plug-and-play parameter-intrinsic watermarking method for neural speech generation. Additionally, DeCoMa presents a dual-channel approach to detect and purify code dataset watermarks, achieving a stable recall of 100% in 14 code watermark detection scenarios.