Artificial intelligence experts are cautioning the public to enhance their cybersecurity measures by using strong passwords and promptly updating software on their devices to combat a new type of cyber-threat known as “AI-driven computer worms.” These advanced worms are capable of launching customized attacks on internet-connected devices, such as laptops, printers, and cameras, as they search for new targets.
Researchers at the University of Toronto, led by Nicolas Papernot, the AI chair at the Canadian Institute for Advanced Research, revealed that publicly available AI models can power a worm that can adapt its attack strategies in real-time while spreading between devices connected to the internet. This groundbreaking research, conducted in collaboration with the Vector Institute, was shared with national science, security, and defense entities before its public release.
Papernot, an associate professor of computer engineering and computer science at U of T, emphasized the importance of updating software promptly and changing passwords regularly during a panel discussion at the university. He stressed the necessity of using multi-factor authentication and ensuring that organizations deploy software patches quickly to mitigate cyber risks.
Unlike traditional computer viruses, worms can spread autonomously from one device to another without human intervention. The AI-driven worm created by the U of T team collects information as it moves across devices, exploiting vulnerabilities and weak points to infiltrate new machines. In contrast to past attacks that followed fixed scripts, these AI-driven worms can design customized attack strategies tailored to each victim device, posing a significant challenge to cybersecurity defenses.
Papernot’s warnings coincide with escalating concerns about the capabilities of AI technology. Recent incidents involving rogue AI agents hacking into online platforms have raised alarms among tech experts about the potential risks of AI systems evading human control. The emergence of sophisticated AI-driven worms underscores the urgent need for enhanced cybersecurity measures to safeguard critical infrastructure and personal data from evolving cyber threats.
The increasing affordability and effectiveness of AI-driven worms make them a formidable challenge for cybersecurity professionals. Samir Chhabra, director general of marketplace framework policy at Innovation, Science and Economic Development Canada, highlighted the lower costs associated with deploying these worms, enabling hackers to target a larger number of victims. This paradigm shift in cyber defense underscores the critical importance of proactive cybersecurity practices to mitigate the growing threats posed by AI-driven attacks.