Social virtual reality (VR) is rapidly transforming how people interact. It creates immersive virtual spaces for social engagement, such as leisure activities, remote work, training and other social events. However, the rise of these platforms has brought significant challenges to user safety. The SAFE-VR (Safety and Awareness For Everyone in Virtual Reality) project aims to develop a comprehensive system to mitigate harassment in social VR by automatically detecting it and providing immediate user support. The first objective will be achieved by analyzing group movement patterns with machine learning methods to detect harassment in social VR. The second objective consists of two parts: 1) dynamic, context-aware personal boundaries that adjust based on the nature of interactions and user preferences, and 2) an intuitive reporting system that automatically captures relevant data when harassment is detected or a report is initiated, easing the process for users. These solutions will be integrated and delivered through the open-source SAFE-VR platform. Perceptual studies will be conducted to evaluate the platform’s efficiency, effectiveness, and its impact on user behavior and sense of safety. An industrial secondment with Virtual Bodyworks, a company specializing in VR for behavioral change in corporate and rehabilitation settings, will provide opportunities to explore real customer needs and tailor the SAFE-VR platform for their commercial VR applications. By delivering adaptable safety mechanisms, this project has the potential to influence future social VR safety standards and benefit developers, users, and policymakers alike.