Executive Summary
A new Android malware strain named RatHat, attributed to China-based threat actors, has emerged with highly advanced capabilities for financial fraud and persistent surveillance. The malware is primarily distributed through deceptive phishing sites, smishing, and malvertising. Unlike typical Android RATs, RatHat employs a multi-tiered architecture including a malicious APK, a privileged Go-based agent, and an FRP reverse-proxy client to maintain deep system access.
Technically, RatHat stands out for its use of Generative AI to autonomously navigate device interfaces by serializing Accessibility trees to XML and requesting tap coordinates from AI models. It further evades detection through 'manifest bombs' and bytecode poisoning, and it bypasses FLAG_SECURE protections by reading raw touch coordinates from /dev/input via an ADB shell. This allows the actor to reconstruct PINs and unlock patterns even when standard accessibility scraping is blocked.
This threat represents a significant escalation in mobile malware tradecraft. By operating out-of-lifecycle daemons that survive app uninstallation and leveraging AI for adaptive UI interaction, RatHat makes traditional signature-based and sandbox-reliant security controls largely ineffective. Affected sectors include global financial services and users of popular payment platforms like Alipay and WeChat.
Key Details
Threat Name
RatHat Android Trojan
Affects
Check Point
Adversary
China-based operators Other Adversaries and Aliases: China-based threat actors; RatHat
MITRE Techniques
Malware/Tools
RatHat, SpiceRAT, BambooToken, Chosen Brick, FRP, liblocal-service.so, libmedia_codec.so, Mantax Otax