The field of image processing and security is moving towards developing innovative solutions to protect image integrity and prevent unauthorized usage. Researchers are exploring new techniques to detect and prevent the use of unauthorized data in training diffusion models, and to ensure the reliability and trustworthiness of medical image reconstruction. Additionally, there is a growing focus on developing robust watermarking schemes and watermark removal methods to protect digital assets. The security of semantic communication systems is also being re-examined, with new threats and vulnerabilities being identified and addressed. Notable papers in this area include HMARK, which proposes a novel multi-bit watermarking scheme, and TokenPure, which introduces a Diffusion Transformer-based framework for effective watermark removal. Rethinking Security in Semantic Communication also stands out for uncovering a fundamental latent-space vulnerability in semantic communication systems.