“Morris II”: The Emergence of Self-Replicating Threats in the Era of Generative Artificial Intelligence

The advancement of generative artificial intelligence (GenAI) has introduced new cybersecurity challenges beyond traditional malware and software vulnerabilities. This paper examines Morris II, a conceptual AI-driven threat based on self-replicating adversarial prompts that exploit the functionality of large language models to propagate, manipulate outputs, and potentially extract sensitive information. Unlike conventional worms that rely on malicious code, Morris II leverages AI interactions and contextual memory to spread through connected systems. The paper analyzes its operational principles, potential attack vectors, and implications for military and critical infrastructure environments. The findings highlight the need for AI-specific security measures, including content filtering, access control, human oversight, and continuous monitoring, to mitigate emerging risks associated with autonomous AI systems.