Cyber Security News

Malicious Android & iOS Apps Downloaded Over 242,000 Times, Stealing Crypto Recovery Keys

A sophisticated malware campaign, dubbed SparkCat, has infiltrated Google Play and Apple’s App Store, marking the first known instance of an optical character recognition (OCR)-based cryptocurrency stealer on iOS.

According to cybersecurity firm Kaspersky, the malware has been downloaded over 242,000 times since its emergence in March 2024.

It targets sensitive cryptocurrency wallet recovery phrases stored in images, posing a significant threat to users across Europe, Asia, and beyond.

Negative user feedback about ComeCome

How SparkCat Operates

SparkCat is embedded within malicious software development kits (SDKs) integrated into seemingly legitimate apps.

On Android, it operates via a Java-based SDK named “Spark,” disguised as an analytics module.

Suspicious SDK being called

For iOS, the malware uses a malicious framework under aliases like “GZIP” or “googleappsdk,” written in Objective-C and obfuscated with HikariLLVM for stealth.

The malware employs Google ML Kit’s OCR technology to scan image galleries for recovery phrases mnemonics critical for accessing cryptocurrency wallets.

These phrases are extracted from screenshots or notes and uploaded to attacker-controlled servers via encrypted channels, including Amazon cloud storage or a Rust-based protocol.

To avoid detection, SparkCat requests gallery access only during specific user actions, such as initiating support chats.

This selective behavior minimizes suspicion while enabling the malware to execute its primary function covertly.

Widespread Impact and Technical Sophistication

The infected apps span various categories, including food delivery services, AI-powered messaging platforms, and crypto-related tools.

Some apps appear legitimate, while others are designed to lure victims.

The campaign’s cross-platform compatibility and use of the Rust programming language a rarity in mobile applications, highlight its technical sophistication.

Kaspersky’s analysis revealed that SparkCat selectively targets users based on keywords in multiple languages, including English, Chinese, Korean, Japanese, and European languages.

To mitigate the risk posed by SparkCat:

  • Uninstall suspicious apps immediately and run antivirus scans.
  • Avoid storing sensitive information like recovery phrases in screenshots or unencrypted formats.
  • Use secure offline storage solutions or hardware wallets for cryptocurrency keys.
  • Regularly update software and avoid downloading apps from unofficial sources.

Google and Apple have been notified about the infected apps; however, some remain available for download.

Users must exercise caution when granting permissions to apps that request access to sensitive data.

Investigate Real-World Malicious Links & Phishing Attacks With Threat Intelligence Lookup - Try for Free

Aman Mishra

Aman Mishra is a Security and privacy Reporter covering various data breach, cyber crime, malware, & vulnerability.

Recent Posts

MobSF Framework Zero-Day Vulnerability Allows Attackers to Trigger DoS in Scan Results

A recently discovered zero-day vulnerability in the Mobile Security Framework (MobSF) has raised alarms in…

14 minutes ago

AnyDesk Flaw Allows Admin Access Through Weaponized Windows Wallpapers

Cybersecurity enthusiasts and IT administrators worldwide are voicing concerns over a newly discovered vulnerability in…

31 minutes ago

DeepSeek’s New Jailbreak Method Reveals Full System Prompt

The Wallarm Security Research Team unveiled a new jailbreak method targeting DeepSeek, a cutting-edge AI…

1 hour ago

Ubuntu Officially Available on the Updated Windows Subsystem for Linux

Ubuntu has announced its availability on Microsoft’s new tar-based Windows Subsystem for Linux (WSL) distribution…

2 hours ago

BADBOX Botnet Surges: Over 190,000 Android Devices Infected, Including LED TVs

The BADBOX botnet, a sophisticated malware operation targeting Android-based devices, has now infected over 192,000…

16 hours ago

Cybercriminals Leveraging AI to Verify Stolen Credit Card Data

Cybercriminals are increasingly leveraging artificial intelligence (AI) agents to validate stolen credit card data, posing…

16 hours ago