While Extended Reality (XR) systems face many of the same cyber threats as traditional computing, the way users are warned in immersive environments requires special care. In XR, a delayed or unclear alert can leave users unknowingly exposed to performance-degrading attacks such as denial-of-service (DoS).
At SUN, we are developing a real-time Intrusion Detection System (IDS) powered by machine learning, designed to identify subtle signs of attack, such as sudden frame drops, sensor overloads, or lag in interactions, even before users even notice them.
Across our ongoing pilot experiments at the University of Greenwich, we are already seeing encouraging results. The IDS shows fast and reliable detection (0.27–1.25 s latency) with no false positives. Participants describe the warnings as clear, timely, and actionable, and most either respond immediately or confirm the alert before acting, indicating strong comprehension and trust in the system.These early results highlight the potential of human-centred cybersecurity in XR: safeguarding users and preserving trust without breaking immersion.