<p>This study examines the impact of generative AI-powered virtual assistants on learning outcomes, user experiences, and interaction behaviors in virtual reality (VR) environments within a museum context. Using a comparative experimental design, the research evaluates two participant groups: one experiencing a VR museum with AI-based virtual assistants and another engaging with a non-AI VR environment. User experience was assessed using the User Experience Scale questionnaire, knowledge acquisition was measured through pre- and post-test evaluations, and open-ended questions captured qualitative feedback. Both groups demonstrated significant improvements in knowledge acquisition, though the integration of AI did not yield statistically significant differences compared to the non-AI group. The AI-assisted group exhibited higher levels of engagement, spending more time exploring exhibits and interacting within the virtual environment. However, participants in the AI-assisted group rated lower in aesthetics and satisfaction, highlighting response delays and inconsistencies as key challenges. These findings underscore both the potential and limitations of integrating AI-based virtual assistants in educational and cultural settings, offering actionable insights for enhancing AI-driven learning experiences in VR while addressing its challenges in educational applications.</p>