AUTO-UPDATED

This new AI-powered worm spreads itself and adapts in real time — here's how to stop it

University of Toronto researchers have developed the first autonomous, AI-powered computer worm capable of learning and adapting its attack strategies in real time to compromise network-connected devices.

Key Points

  • The proof-of-concept malware utilizes publicly available large language models to identify and exploit vulnerabilities without human intervention.
  • Unlike traditional static malware, this autonomous worm can spread across networks independently and repurpose infected device processing power to sustain its operations.
  • Researchers developed the tool in a controlled academic setting to proactively identify security risks before malicious actors can weaponize similar AI technologies.
  • Experts recommend securing devices by changing default passwords, utilizing passkeys, enabling multi-factor authentication, and installing software updates immediately.
  • The study highlights that smart home appliances, routers, and cameras are particularly susceptible to these adaptive threats if not properly isolated on guest networks.

Why it Matters

This development signals a shift toward a new era of cybersecurity where autonomous AI agents can execute sophisticated, self-evolving attacks at minimal cost. Understanding these vulnerabilities is essential for strengthening digital infrastructure and protecting personal data against increasingly intelligent automated threats.
MakeUseOf Published by Sara Heritage
Read original