ESET Discovers First Android Malware Exploiting Generative AI
PromptSpy marks the first known use of generative AI in Android malware, giving attackers the ability to adapt to different devices and OS versions through dynamic UI analysis.
Reporting from 1 sources: ASCII.jp.
ESET researchers have identified PromptSpy, the first Android malware to use generative AI. It exploits Google Gemini to analyze screens and issue instructions for persistence, preventing forced termination. The malware deploys a VNC module for remote device control and targets users in Argentina, with development clues pointing to a Chinese-language environment.
ESET researchers have uncovered PromptSpy, the first Android malware family that leverages generative AI in its execution flow. The malware uses Google Gemini to interpret on-screen elements on compromised devices and generate step-by-step instructions that keep the malicious app pinned in the recent apps list, preventing forced termination by the system or swipe deletion. While the AI component is limited to persistence, it allows the malware to adapt to different device layouts and OS versions.
PromptSpy's primary function is to deploy a built-in VNC module, giving attackers remote control over the victim's device. It also exploits accessibility services to block uninstallation, capture lock screen data, and record video. Communication with the command-and-control server uses the VNC protocol with AES encryption. ESET reports that the campaign appears financially motivated and primarily targets users in Argentina, with development clues suggesting a Chinese-language environment. The malware is distributed through dedicated websites, not Google Play, and Google Play Protect automatically blocks known versions.
Synthesized by Yomimono from the 1 cited source below, including Japanese-language reporting where cited, then editorially reviewed before publishing.