The field of memory security and embedded systems is witnessing significant advancements, with a focus on developing innovative solutions to mitigate memory-related vulnerabilities and improve the security and performance of embedded devices. Researchers are exploring new approaches to memory management, such as cluster-based memory allocation, to reduce tag collisions and enhance the effectiveness of tag-based sanitizers. Additionally, there is a growing interest in leveraging artificial intelligence and machine learning techniques to generate and validate secure embedded firmware, as well as to develop dynamic vulnerability patching frameworks for heterogeneous embedded systems. Empirical evaluations of memory-erasure protocols are also being conducted to assess their feasibility, performance, and security on real devices. Noteworthy papers in this area include: Securing LLM-Generated Embedded Firmware through AI Agent-Driven Validation and Patching, which proposes a three-phase methodology for generating and validating secure firmware. Dynamic Vulnerability Patching for Heterogeneous Embedded Systems Using Stack Frame Reconstruction, which introduces a hot patching framework for embedded devices with limited computational power and memory.