AI TwinGuard System Rapidly Detects 5G Cyber-Attacks

• •TwinGuard can detect 5G cyber-attacks in under 100 milliseconds.
• •The system uses a real-time digital twin and reinforcement learning AI.
• •Future 6G networks will need security systems that learn behavioral patterns.

Researchers at Surrey University have developed TwinGuard, an AI defense system that can detect and neutralize sophisticated 5G cyber-attacks in under 100 milliseconds. The system utilizes a real-time digital twin of mobile networks, which updates every few milliseconds, and is paired with reinforcement learning AI to anticipate and respond to suspicious behaviors. TwinGuard was tested in two realistic 5G environments, successfully blocking attacks such as handover flooding and E2 subscription flooding. The increasing openness of 5G networks presents more opportunities for cyber attackers, making advanced detection systems essential. The research was presented at the 2025 IEEE International Conference on Trust, Security and Privacy in Computing and Communications. Future plans include expanding TwinGuard to larger multi-cell environments for deployment in upcoming 6G systems. This advancement is crucial as 6G networks are expected to require adaptive security measures against evolving threats.