Collaborative immersive systems use augmented, virtual, and mixed reality to merge physical and digital worlds, allowing multiple users to share and interact within the same three-dimensional space in real time through natural actions such as gaze, gestures, and spatial manipulation. By creating a sense of shared presence, these systems enable new forms of collaboration in critical domains such as healthcare, education, and workforce training. However, unlike web or mobile platforms where users interact independently through a screen, each user's actions in these systems are sensed and directly affect what every other user sees and experiences. This introduces security and privacy risks that are fundamentally different from those in traditional computing, as a malicious participant can manipulate what others see and experience, impersonate trusted users, or exploit the sensitive data that these systems continuously capture about users' bodies, movements, and surroundings, risks that are amplified as these systems integrate AI-driven features. Existing security approaches cannot address these risks because they do not account for the multi-user, embodied nature of immersive interaction. The project's novelties are a principled framework that formally models the multi-layered interactions unique to collaborative immersive environments and provides integrated mechanisms for trust establishment among participants, secure provenance of shared interactions, and adaptive security