PupkinStealer : A .NET-Based Info-Stealer

PupkinStealer : A .NET-Based Info-Stealer
PupkinStealer is a .NET-based malware designed to steal browser credentials, desktop files, Telegram sessions, and Discord tokens. It exfiltrates data via Telegram Bot API, using simple, stealthy methods without persistence. (Affected: Windows users, messaging platforms, enterprises)

Keypoints :

  • PupkinStealer targets Chromium-based browser credentials, desktop files (.pdf, .txt, .sql, .jpg, .png), Telegram sessions, Discord tokens, and desktop screenshots.
  • It decrypts browser passwords by extracting keys from Local State files using Windows Data Protection API.
  • The malware copies Telegram’s tdata folder to hijack active sessions without reauthentication.
  • Discord tokens are harvested from leveldb folders using regex to enable unauthorized access.
  • A screenshot of the desktop at 1920×1080 resolution is captured and saved locally before exfiltration.
  • All stolen data is compressed into a ZIP archive embedding victim metadata (username, IP, SID) for attacker tracking.
  • The data is exfiltrated through a Telegram Bot API channel, leveraging Telegram’s anonymity for stealth.
  • PupkinStealer shows no persistence or anti-analysis features, relying on low-profile execution for evasion.
  • Attribution links the malware to an actor alias “Ardent,” with indications of Russian language usage.
  • Recommendations include endpoint security, network monitoring, application whitelisting, and user awareness training to mitigate risks.

MITRE Techniques :

  • Command and Scripting Interpreter (T1059) – Uses .NET runtime to execute malware code and launch asynchronous data harvesting tasks.
  • Malicious File (T1204.002) – Deploys an unsigned executable that runs without user consent to harvest data.
  • Compression (T1027.015) – Compresses stolen data into a ZIP archive with maximum compression to reduce detection surface.
  • Registry Run Keys / Startup Folder (T1547.001) – (Reported in MITRE table, though PupkinStealer lacks persistence, this technique is generally related to persistence.)
  • Credentials from Web Browsers (T1555.003) – Extracts and decrypts saved Chromium browser credentials using DPAPI keys.
  • Steal Application Access Token (T1528) – Harvests Discord authentication tokens from local storage to impersonate users.
  • System Information Discovery (T1082) – Gathers system metadata (username, IP, SID) embedded in exfiltration data.
  • Data from Local System (T1005) – Collects files from desktop with targeted extensions for sensitive data.
  • Screen Capture (T1113) – Takes screenshots of the victim’s desktop for additional information gathering.
  • Local Data Staging (T1074.001) – Stores collected files in temporary directories before compression and exfiltration.
  • Exfiltration Over C2 Channel (T1041) – Sends stolen information through Telegram Bot API to attacker-controlled channel.
  • Exfiltration to Cloud Storage (T1567.002) – Uses Telegram cloud infrastructure to host and receive exfiltrated data.

Indicator of Compromise :

  • The article lists file hashes for PupkinStealer.exe, including MD5 (fc99a7ef8d7a2028ce73bf42d3a95bce) and SHA-256 (9309003c245f94ba4ee52098dadbaa0d0a4d83b423d76c1bfc082a1c29e0b95f).
  • Telegram Bot API URLs and tokens used for data exfiltration, e.g., API endpoint with token “8013735771:AAE_UrTgQsAmiAsXeDN6mehD_fo3vEg-kCM.”
  • File paths of harvested data on victims’ machines: browser credentials (%APPDATA%Temp[Username]GrabbersBrowserpasswords.txt), Telegram sessions (%APPDATA%Temp[Username]GrabbersTelegramSession*), Discord tokens (%APPDATA%Temp[Username]GrabbersDiscordTokens.txt), screenshots (%APPDATA%Temp[Username]GrabbersScreenshotScreen.jpg), collected desktop files (%APPDATA%Temp[Username]DesktopFiles*), and the final zipped archive (%APPDATA%Temp[Username][Username]@ardent.zip).
  • These IOCs provide visibility into file artifacts, network exfiltration endpoints, and token theft enabling detection and response.



Read more: https://www.cyfirma.com/research/pupkinstealer-a-net-based-info-stealer/

Views: 36