Researchers at the University of Toronto say their project highlights the risk of open-weight AI models being weaponized. The worm was able to spread over a network with no human guidance.
Some results have been hidden because they may be inaccessible to you
Show inaccessible results