Abstract:
The convergence of virtual reality (VR) and the rapidly evolving Metaverse is transforming
how we interact with technology, creating unprecedented opportunities for immersive
experiences and novel forms of social interaction. This exciting new frontier, however, is
marred by a significant challenge: the absence of comprehensive ethical frameworks and
robust governance mechanisms within these digital spaces. This deficiency fosters a
climate in which unethical behavior, including the alarming prevalence of hate speech, can
flourish unchecked. While initiatives like Extended Reality Safety Initiative (XRSI) and
MetaHate attempt to address these concerns, practical implementation within real-world
virtual environments remains a substantial hurdle, especially for developers. This research
proposes a novel ethical speech recognition system designed to detect unethical language
within VR and the metaverse. This system is built on Unreal Engine (UE), utilises the
Runtime Speech Recognizer (RSR) plugin, incorporating a dynamic dictionary of unethical
phrases and real-time detection. Initial evaluations show the system's potential to identify
unethical language in VR interactions, showcasing its efficacy in a multiplayer VR
environment. This research, therefore, contributes significantly to the ongoing efforts to
build ethical and inclusive immersive technologies, paving the way for safer and more
responsible interactions in virtual environments
