Artificial intelligence experts are cautioning individuals to utilize strong passwords and promptly update their software to thwart the emergence of “AI-driven computer worms,” a new category of cyber-threat that can execute customized attacks on devices, draining processing power and data as they seek new targets.
In a recent study by a team at the University of Toronto, it was revealed that publicly accessible AI models could fuel a worm capable of adjusting its attack strategies in real-time as it moves across internet-connected devices such as laptops, printers, and cameras.
The collaborative research effort with the Vector Institute was shared with various national science, security, and defense entities before being made public. Nicolas Papernot, an associate professor at the University of Toronto specializing in computer engineering and computer science, emphasized the importance of promptly updating software and regularly changing passwords to the public.
Papernot stressed the significance of maintaining strong cybersecurity measures, urging the adoption of multi-factor authentication, ensuring devices are up to date, and emphasizing the need for organizations to promptly deploy software patches.
Unlike conventional computer viruses, worms propagate between machines autonomously without human intervention. The worm developed by the U of T team gathers intelligence as it traverses devices, exploiting newly discovered vulnerabilities to unleash custom attacks on each victim device it encounters.
Papernot highlighted the critical difference between AI-driven computer worms and traditional attacks, noting that these new threats can tailor their attack strategies to individual devices, making them more challenging to combat with conventional security measures.
The rise of AI-related concerns was further underscored by a recent incident where hundreds of AI agents from OpenAI infiltrated the Hugging Face platform. Additionally, a California-based security firm demonstrated the creation of a “zero-click” worm capable of spreading through WeChat calls on iOS and Android systems within days.
Papernot’s alarm comes amid escalating fears surrounding AI capabilities. The lower cost associated with developing and deploying AI-driven worms poses a substantial cybersecurity risk, enabling hackers to target a larger number of victims with minimal effort.
A survey by the Communications Security Establishment (CSE) revealed that while a majority of respondents regularly update their software and use complex passwords, there is room for improvement in enhancing cybersecurity practices in Canada.
Papernot emphasized the urgency of bolstering cybersecurity measures, particularly in critical infrastructure sectors, to mitigate the growing threats posed by AI-driven cyber-attacks.