This study presents a framework for integrating agents of generative artificial intelligence (AI) within immersive virtual reality (VR) environments to support high school anatomy education. While traditional tools offer common visual aids or textual descriptions of human anatomy, our approach employs a Unity-based 3D VR simulation, combined with a large language model (LLM)-powered AI instructor, to facilitate interactive, personalized learning. The system enables students to construct and deconstruct a highly comprehensive virtual human body layer by layer, engaging with both internal and external anatomical structures through Mechdyne’s ARC System and various VR headset models. This study also explores the possibility of introducing an AI agent with real-time natural language processing and text-to-speech capabilities, allowing it to respond conversationally to student inquiries and adapt its guidance based on prior knowledge, demographic data, and learning assessments. Another exploration in this paper is the data architecture behind the AI agent, which should record student interactions, question types, and knowledge gaps. This information will be used to fine-tune AI responses via a reinforcement-learning-based personalization layer. The entire system should eventually be evaluated through a mixed-methods study that includes usability testing, pre- and post-knowledge assessments, and student feedback surveys.