CybersecurityAI Research14 min reading time

Import AI 467: Self-sustaining AI viruses; pacing AI progress; confusion about AI and creativity

Import AI
Read full post
Researchers from the University of Toronto, Vector Institute, University of Cambridge, and ServiceNow developed a prototype AI-driven computer virus that uses large language models (LLMs) running on compromised GPUs to autonomously identify vulnerabilities and spread itself. This proof-of-concept virus operates without relying on external APIs and demonstrates the feasibility of self-sustaining AI cyber threats.

More in Cybersecurity

Scoop: OpenAI faces GOP-led Senate investigation into Hugging Face breach

Covered by 3 sources
Cybersecurity6 min read

Anthropic Discloses Fourth AI Hacking Incident Involving Claude Opus 4.6

Covered by 2 sources
Cybersecurity5 min read

Every AI Incident Has Two Timelines. We Default To One

Forbes