Back to News
NewsAi

Researchers Warn AI Agents Could Be Exploited as Botnets via Hallucinations

Researchers warn that AI hallucinations can be exploited to create botnets, posing significant cybersecurity risks.

Mercation Admin
July 9, 2026 1 min read
Researchers Warn AI Agents Could Be Exploited as Botnets via Hallucinations

New Attack Method Exploits AI Hallucinations

Recent findings from a collaboration between Tel Aviv University, Technion, and Intuit have unveiled a novel attack technique termed "Adversarial HalluSquatting." This method capitalizes on AI models' propensity to generate hallucinations, leading to potential security breaches.

The research indicates that AI agents can be misled into trusting fake repositories or tools that contain malicious instructions, posing a significant threat in the evolving landscape of AI applications. The study, titled "Beware of Agentic Botnets: Scalable Untargeted Promptware Attacks via Universal and Transferable Adversarial HalluSquatting," outlines how AI-generated fictitious links can be weaponized.

As AI assistants transition from merely answering questions to executing tasks, including writing code and accessing files, the risk of these agents acting on unverified information increases. The researchers found that during tests involving AI coding assistants, such as GitHub Copilot and OpenClaw, hallucinated resources were generated with alarming frequencies—up to 85% during repository cloning and 100% during skill installations.

HalluSquatting resembles typosquatting, where attackers register domain names that mimic legitimate ones. However, it specifically targets AI errors rather than human mistakes. This exploitation could lead to the creation of AI-powered botnets, which are collections of compromised devices used for various cyberattacks, including denial-of-service and ransomware operations.

Furthermore, the researchers emphasize that the implications of these findings could be extensive, potentially leading to financial, privacy, and security risks associated with AI interactions. They noted ongoing research into various promptware attacks that threaten real-world systems, highlighting the urgent need for enhanced security measures in AI applications.