New RatHat Android Malware Uses AI to Control Devices

Zimperium zLabs researchers have analyzed a new Android malware family called RatHat that uses an AI-powered subsystem to help operators remotely navigate compromised devices. The researchers believe it may be linked to threat actors from China after finding Chinese-language prompts for a large language model.

RatHat is distributed through malvertising, SMS messages, and phishing sites that promote APK downloads outside Google Play. It abuses Android Accessibility permissions to perform privileged actions, enables Developer Options and Wireless Debugging, and creates a local shell-level execution context without an external computer. Similar mechanisms have appeared in ToxicPanda and RedHook.

A Go-based agent, liblocal-service.so, runs commands with ADB shell privileges, bypasses battery restrictions, and maintains persistence. The malware can restore the agent if it is removed or stopped, while the agent can restore the malware. Another agent, libmedia_codec.so, establishes a persistent FRP reverse-proxy tunnel to the attacker.

RatHat displays HTML overlays targeting banking and cryptocurrency apps to capture credentials. It can also intercept SMS messages and notifications, including one-time passwords, extract browser URLs, and capture lock-screen PINs, passwords, and unlock patterns.

Zimperium says the malware serializes the live Android Accessibility tree into XML and sends it to an unnamed popular AI assistant. Researchers also observed uninstall interception, a fake Google Play error overlay, APK tampering, a bloated 61MB manifest, and invalid DEX pseudo instructions designed to hinder analysis.