6/8/26 A new paper from the University of Toronto researches the capabilities of AI-powered worms – and it comes close to the dystopian worms of Science Fiction over the past 25 years. The AI-powered worms prototyped can autonomously adapt their attack strategy as they move through a network.
Unlike traditional worms that exploit a single vulnerability, these AI-enabled worms can identify weaknesses in each device they encounter, tailor their approach, and continue spreading with little to no human intervention. Researchers demonstrated the concept using publicly available AI models in an isolated lab environment.
For security leaders, attack automation is becoming more adaptive and scalable increasing risk to organizations. It is critical to address this threat in your environment through security risk mitigations, which include security hygiene, resilient defensive strategies, and vulnerability management, identity security, segmentation, and rapid patching.
Research like this suggests we may be approaching a point where AI not only accelerates attacker productivity but also enables malware to dynamically reason and adjust during an attack.
https://www.nytimes.com/2026/06/02/technology/scientists-find-way-to-supercharge-dangerous-computer-worms-with-ai.html