Molecular Communication (MC) is an emerging field of research focused on understanding how cells in the human body communicate and exploring potential medical applications. In theoretical analysis, the goal is to investigate cellular communication mechanisms and develop nanomachine-assisted therapies to combat diseases. Since cells transmit information by releasing molecules at varying intensities, this process is commonly modeled using Poisson channels. In our study, we consider a discrete-time Poisson channel (DTPC). MC is often event-driven, making traditional Shannon communication an unsuitable performance metric. Instead, we adopt the identification framework introduced by Ahlswede and Dueck. In this approach, the receiver is only concerned with detecting whether a specific message of interest has been transmitted. Unlike Shannon transmission codes, the size of identification (ID) codes for a discrete memoryless channel (DMC) increases doubly exponentially with blocklength when using randomized encoding. This remarkable property makes the ID paradigm significantly more efficient than classical Shannon transmission in terms of energy consumption and hardware requirements. Another critical aspect of MC, influenced by the concept of the Internet of Bio-NanoThings, is security. In-body communication must be protected against potential eavesdroppers. To address this, we first analyze the DTPC for randomized identification (RI) and then extend our study to secure randomized identification (SRI). We derive capacity formulas for both RI and SRI, providing a comprehensive understanding of their performance and security implications.