Self-poisoning Network to Prevent Reconnaissance by Generative Artificial Intelligence

The incorporation of generative AI in reconnaissance significantly streamlines the process for hackers, facilitating the analysis of data and extraction of valuable insights. In this paper, we have explained how we can reinforce network defense by preventing the use of generative AI in such attacks. We have shown how using Scapy, a Python-based program, and a couple of Python commands, we can transmit crafted packets containing confusing network metadata, as well as payloads which could contain AI-specific malware, between specific devices, thus self-poisoning the network. This can provide multiple layers of security to the network. To demonstrate the process, we have used the said tools in Kali Linux environment. We have also outlined a self-poisoning algorithm for network defense system developed through integrating AI poisoning technique, with generative adversarial network.