When one individual or group uses a computer network to attack another, it is called a cyber-assault. When an internet resource's accessibility, privacy, or integrity are compromised in more than one way, it is considered an incursion. When an intrusion detector identifies suspicious behavior on a network or system - for example, a breach of security policies, it will notify the management station. It is not practical to identify threats in real-time due to the massive amounts of data sent by cyber systems. This research aims to identify cyber threats by using artificial neural networks. It may be possible to detect fraud by monitoring user behavior for irregularities. How computers can identify new weaknesses and react to fix or remove them constitutes one of the greatest mysteries in cyber security. To prevent cyber-attacks, new cyber security measures are needed to identify malicious nodes before they communicate. To improve system performance and social media, this research suggests the Swift Self-Replication Model with Self-Repair Mode Enabled Distributed Network for Intrusion Detection (SSRM-SRM-DNIDS). The self-replication model equips compromised systems with the necessary code to fix themselves, and our method efficiently identifies intrusions in dispersed environments. The suggested approach has Malware Detection Rate of 97.6% accuracy rate. The suggested method has a 98.4% success rate with regard to intrusion detection and self-replication. We are seeing a decline in false positives. A computerized approach is required to eliminate false positives during real time because most solutions work offline. © 2025 Elsevier B.V., All rights reserved.