As reported by SecurityAffairs, Zimperium's zLabs team has documented a major evolution of the ToxicPanda Android banking trojan, now targeting 349 financial institutions across 16 countries with 167 remote commands. While the scale expansion is noteworthy, the technical innovations — particularly the abuse of Android's Wireless Debugging framework — deserve close attention from mobile security teams.

Threat Alert: As reported by SecurityAffairs, Zimperium's zLabs team has documented a major evolution of the ToxicPanda Android banking trojan, now targeting 349 financial institutions across 16 countries with 167 remote commands.

Scale Meets Sophistication

The 22-fold increase in targeted apps (from 16 to 349) signals a shift from a regional nuisance to a globally distributed banking fraud platform. But the more concerning development is operational maturity. Several command structures that Cleafy previously identified as stubs or unimplemented in earlier versions are now fully functional. This tells us the developers are actively iterating based on prior research exposure — a pattern more commonly associated with commercially distributed malware-as-a-service families than with boutique banking trojans.

The combination of Accessibility Service abuse, Google Play Protect suppression via VPN permissions, and now Wireless Debugging escalation creates a layered defense-evasion stack that most consumer-grade mobile security products will struggle to catch in real time.

Wireless Debugging: The Underappreciated Attack Surface

The most technically significant development is the abuse of Android's Wireless Debugging feature. Introduced in Android 11, Wireless Debugging allows ADB connections over Wi-Fi without a physical USB cable. While convenient for developers, it exposes a broader attack surface because it can be enabled programmatically under certain conditions and does not require the same user consent friction as USB debugging.

By leveraging Accessibility Service permissions to navigate settings menus and enable Wireless Debugging silently, ToxicPanda 2.0 can establish a persistent, higher-privilege communication channel that operates outside the visibility of most mobile device management (MDM) and endpoint detection tools. This is a meaningful escalation because:

Wireless Debugging: The Underappreciated Attack Surface
It bypasses network-level monitoring — ADB over Wi-Fi operates on local network segments that corporate monitoring may not inspect.
It enables deeper device control — ADB-level access permits file system exploration, app installation, and log access beyond what Accessibility Service alone provides.
It persists across reboots — Once debugging is enabled and paired, the connection can be re-established without repeated user interaction.

Accessibility Service Abuse Remains Android's Achilles Heel

ToxicPanda 2.0's core credential theft mechanism still relies on the Accessibility Service — the same attack vector that has powered Android banking trojans for years. Despite Google's incremental restrictions on which apps can request accessibility permissions, the fundamental design tension remains: the service must be powerful enough for legitimate assistive technology, which makes it equally powerful for malicious screen-scraping and UI automation.

For financial institutions, this means server-side fraud detection remains the most reliable layer of defense, as client-side controls on consumer devices are inherently unreliable against accessibility-based attacks.

Who Is Most at Risk

  • Retail banking customers in the 16 targeted countries, particularly those who sideload apps or install from third-party stores.
  • BYOD-enabled enterprises where personal devices access corporate banking or financial applications.
  • Fintech and neobank customers who may be disproportionately targeted due to lighter fraud controls compared to legacy banks.

Shield53 Recommendations

For Financial Institutions

  • Enhance server-side behavioral fraud analytics to detect anomalous session patterns indicative of remote-controlled devices — unexpected navigation sequences, inhuman interaction timing, or simultaneous multi-app access.
  • Implement app integrity attestation (e.g., Google Play Integrity API) to detect devices running with debugging enabled or accessibility services associated with untrusted packages.
  • Deploy or update mobile threat defense solutions that specifically flag Wireless Debugging activation on non-developer profiles.

For Enterprise Security Teams

  • Update MDM policies to disable Wireless Debugging on managed Android devices and alert on any attempt to enable it.
  • Restrict Accessibility Service permissions to a vetted allowlist of approved applications.
  • Block sideloading and third-party app stores on corporate-managed devices; for BYOD, implement conditional access policies that deny corporate resources to devices with untrusted accessibility services enabled.

For End Users

  • Never install Android apps from outside Google Play unless you can verify the publisher's authenticity.
  • Review which apps have Accessibility Service permissions under Settings > Accessibility and remove any unrecognized entries immediately.
  • Ensure Wireless Debugging is disabled (Settings > Developer Options > Wireless Debugging) unless actively developing.
  • Keep Google Play Protect enabled and do not grant VPN permissions to apps that have no legitimate reason to request them.

The trajectory of ToxicPanda — from a limited European banking trojan to a 349-target global platform with novel privilege escalation techniques — illustrates that the Android malware ecosystem is maturing rapidly. Defenders should treat the Wireless Debugging attack vector as an emerging trend likely to be adopted by additional families in the coming months, not as a one-off capability.