"New ConfusedPilot Attack Targets AI Systems with Data Poisoning"

Researchers from the University of Texas at Austin's SPARK Lab have identified "ConfusedPilot," a new cyberattack that targets Retrieval-Augmented Generation (RAG)-based Artificial Intelligence (AI) systems such as Microsoft 365 Copilot. Professor Mohit Tiwari, CEO of Symmetry Systems, led the team that discovered how attackers could manipulate AI-generated responses through the introduction of malicious content into documents referenced by the AI. This method could result in misinformation and flawed decision-making by organizations. The potential for widespread disruption is high, with 65 percent of Fortune 500 companies adopting or planning to use RAG-based systems. This article continues to discuss the new ConfusedPilot cyberattack method.

Infosecurity Magazine reports "New ConfusedPilot Attack Targets AI Systems with Data Poisoning"

Submitted by grigby1