Abstract

This chapter discusses the role of human-AI partnership within the enterprise metaverse and how AI hallucinations endanger trust, emotional congruency and productivity. The more immersive virtual environments transform organizational work, the more AI agents help in analyses, decision support, and training, but, because of their probabilistic nature, they can give confident and false results. The chapter introduces the TrustEmotion Productivity (TEP) Framework to demonstrate that safe collaboration relies on calibrated trust, emotional intelligence, and design that is qualified by productivity orientation. It also emphasizes the intensification of hallucinations within immersive, real-time, and multimodal environments and analyzes technical and organizational measures like retrieval-augmented generation, confidence calibration, transparency, expert review and governance protocols. It illustrates the operational risk of hallucinations and best practices of constructing credible, ethical, and human-focused AI systems to support metaverse-enabled businesses.