Search This Blog

Powered by Blogger.

Blog Archive

Labels

Footer About

Footer About

Labels

Showing posts with label .desktop file abuse. Show all posts

Google Introduces AI-Powered Ransomware Detection in Drive for Desktop

 

Ransomware continues to be a growing cyber threat, capable of crippling businesses and disrupting personal lives. Losing access to vital files — from cherished family photos to financial records — can have devastating consequences. To tackle this, Google is introducing an AI-powered ransomware detection system for Drive for Desktop, designed to identify threats early and prevent large-scale data loss.

According to Google’s blog post, this new security layer for macOS and Windows continuously monitors for abnormal behavior, such as mass file encryption or corruption — common indicators of a ransomware attack. Unlike traditional antivirus tools that scan for malicious code, Google’s AI model focuses on how files change. When it detects unusual activity, even across a few files, it immediately halts syncing between the user’s device and the cloud. This pause prevents infected files from overwriting safe versions in Google Drive.

Once potential ransomware activity is detected, users receive desktop and email alerts and can access a new recovery interface within Drive. This interface allows them to restore their files to a clean, pre-attack state.

Ransomware remains a significant cybersecurity issue. In 2024, Mandiant reported that ransomware accounted for 21% of all intrusions, with an average cost per incident exceeding $5 million. Critical industries such as healthcare, education, retail, manufacturing, and government are particularly at risk. Google’s approach focuses on a crucial middle ground — between traditional antivirus prevention and post-attack recovery — where AI-driven early intervention can make a major difference.

Google emphasizes that this feature isn’t meant to replace antivirus or endpoint detection tools but to act as an additional safeguard. The system prioritizes commonly targeted file types like Office documents and PDFs, while native Google Docs and Sheets already benefit from built-in protection. Importantly, Google notes that it does not collect user data to train its AI models without explicit consent.

The AI ransomware detection feature is currently rolling out in open beta and will be available at no extra cost for most Google Workspace commercial customers. Individual users will also have access to file recovery tools for free. However, there’s no confirmation yet on whether similar protections will extend to Google Cloud Storage for enterprise users.

APT36 Exploits Linux .desktop Files for Espionage Malware in Ongoing Cyber Attacks

 


The Pakistani threat group APT36 has launched new cyber-espionage attacks targeting India’s government and defense sectors by abusing Linux .desktop files to deploy malware.

According to recent reports from CYFIRMA and CloudSEK, the campaign—first detected on August 1, 2025—is still active. Researchers highlight that this activity focuses on data theft, long-term surveillance, and persistent backdoor access. Notably, APT36 has a history of using .desktop files in espionage operations across South Asia.
Abuse of Linux Desktop Files

Victims receive phishing emails containing ZIP archives with a disguised .desktop file masquerading as a PDF. Once opened, the file triggers a hidden bash command that fetches a hex-encoded payload from an attacker-controlled server or Google Drive, writes it into /tmp/, makes it executable with chmod +x, and launches it in the background.

To avoid suspicion, the malware also opens Firefox to display a decoy PDF hosted online. Attackers manipulated fields like Terminal=false to hide terminal windows and X-GNOME-Autostart-enabled=true for persistence at every login.

While .desktop files are typically harmless text-based launchers defining icons and commands, APT36 weaponized them as malware droppers and persistence mechanisms—a method similar to how Windows LNK shortcuts are exploited.

The dropped malware is a Go-based ELF executable with espionage capabilities. Despite obfuscation, researchers confirmed it can:
  • Remain hidden,
  • Achieve persistence via cron jobs and systemd services,
  • Establish C2 communication through a bi-directional WebSocket channel for remote command execution and data exfiltration.
Both cybersecurity firms conclude that APT36 is evolving its tactics, becoming increasingly evasive, stealthy, and sophisticated, making detection on Linux environments difficult since .desktop abuse is rarely monitored by security tools.