Researchers from the University of Toronto and the University of Cambridge have developed a novel AI-powered computer worm capable of autonomously generating tailored attack strategies for each targeted device. This new threat, detailed in a preprint paper, utilizes an unnamed open-source LLM to exploit common network vulnerabilities like reused passwords across Linux, Windows, and IoT devices in a controlled environment. Unlike traditional worms that exploit a single vulnerability, this AI worm can dynamically detect and adapt to unique security flaws on each device, making it significantly more dangerous. While its current propagation rate is slower, researchers warn that advancements in device efficiency and AI models will accelerate its spread, posing a substantial risk to internet-connected machines as both data repositories and attack launchpads.