
A team from the University of Toronto built an AI-powered worm using open-source AI models that spreads autonomously across a test network. The worm tailors attacks to specific vulnerabilities and operates without human intervention in a secure closed environment. It targets devices running Linux, Windows, and IoT systems with common corporate flaws like reused passwords. The research was published as a preprint by the University of Toronto and University of Cambridge.
Tracked by ByteBrief