Nobody needs Mythos or 0-days to build a chaos-causing computer worm – free open source models work just fine

Researchers at the University of Toronto built and tested a proof-of-concept LLM-driven autonomous worm in an isolated 33-host "FakeCorp" network; the prototype autonomously identified and weaponized known (and newly published) vulnerabilities and misconfigurations to escalate privileges and propagate, successfully exploiting a large portion of the seeded hosts but operating slowly, without concealment, and in an environment lacking endpoint defenses. The authors withheld some methodological details and the model, did not publicly release code, and coordinated disclosure with relevant agencies while warning that such low-cost models broaden the threat surface.