Skip to content
NewA calmer, faster ZeroHour
StoryMalwarehigh

Fake ChatGPT and Claude apps sideload Yoba malware via signed binaries

7 sources · 9 articlesFirst reported Updated ()
What's new

First merged summary.

Merged summary · x-ai/grok-4.7 · rewritten as coverage arrivesSummaries by Vani AI

Check Point says fake GitHub apps pair signed OpenAI or Anthropic installers with a DLL that has delivered an XMRig miner.

Check Point Research describes a campaign it calls Yoba that uses fraudulent GitHub projects posing as desktop apps named GPT-5.6 Sol, GPT-6 Astra, and Claude Opus. ZIP releases pair genuine signed OpenAI or Anthropic Setup.exe files with a malicious winmm.dll, YobaDropper, which sideloads and installs YobaLoader. OpenAI, Anthropic, ChatGPT, and Claude were not compromised. The loader uses Telegram, a generated domain, DNS-over-HTTPS, and Catbox; an observed payload was XMRig, including version 6.26.0 over Tor and injected into svchost.exe, though the loader can also run other executables or PowerShell. Check Point cites indicators suggesting Russian origins without proof.

Key facts
  • Malicious ZIPs began August 30, 2026; further uploads August 31 and September 4.
  • YobaDropper persists as a Local System scheduled task using Windows Media Player.
  • The dropper skips several CIS keyboard layouts but not Ukrainian.
  • An earlier June 2026 variant impersonated Cursor.exe.
  • Commands are Ed25519-verified and AES-decrypted before execution.
Analyze this story with Vani AIImpact, who is affected and what defenders should do, in a new chat

Timeline

dated events in the reporting · oldest first
  1. Jul 16, 2026First fake GitHub repository in the Yoba campaign · Check Point Research
  2. Aug 30, 2026Malicious ZIP releases began · Check Point Research
  3. Aug 31, 2026Additional malicious release uploaded · CSO Online
  4. Sep 4, 2026Further malicious release uploaded · CSO Online
  5. Oct 27, 2026Check Point publicly described the Yoba campaign · Infosecurity Magazine

What each outlet adds

beyond the shared facts
  • Check Point ResearchFirst repository dated July 16; commit offsets and Russian-origin indicators.
  • Cyber Security NewsTelegram image bytes form a dedyn.io domain; Ed25519 and AES.
  • GBHackersObserved XMRig 6.26.0; vendors were not compromised.
  • CSO OnlineRelease dates through September 4; about a 12-day wait unless launched by Setup.exe.
  • The Hacker NewsHost-name hash check; XMRig injected into svchost.exe.
Vendors
OpenAIAnthropic
Products
ChatGPTClaudeCursorClaude CodeChatGPT DesktopCodexWindsurfGitHub CopilotGemini CLIJetBrains AI AssistantWindows Media PlayerOpenAI Codex
Malware
YobaYobaDropperYobaLoaderXMRigBeagleAokigaharaAokiKamidereStygian
Organizations
Check Point ResearchCheck PointQiAnXinSophosSymantec Threat Hunter TeamCheck Point Software
Countries
Russia
AI models
GPT-5.6 SolGPT-6 AstraClaude Opus

Coverage

oldest first · one row per article · UTC
  1. · 4d ago GBHackersFirst report

    Aokigahara infostealer collects AI coding-tool tokens and cloud credentials

    Check Point says it searches local app storage rather than exploiting a flaw in the targeted AI assistants.

  2. · 3d ago Cyber Security News

    Aokigahara infostealer steals ChatGPT, Claude, Codex, and Cursor logins

    Check Point says the Windows stealer also grabs cloud, SSH, and browser secrets, not via flaws in those AI apps.

  3. · 3d ago Check Point Research

    Fake ChatGPT and Claude GitHub apps sideload Yoba malware

    Check Point says the chain uses signed OpenAI and Anthropic binaries plus Telegram, Tor, Catbox and DNS over HTTPS.

  4. · 3d ago Cyber Security News

    Yoba campaign uses signed ChatGPT and Claude installers to sideload malware

    Fake GitHub ZIPs pair signed OpenAI or Anthropic Setup.exe with a rogue DLL that loads an XMRig miner via Telegram and DNS.

  5. · 3d ago Dark Reading

    Aokigahara stealer takes developer keyrings and AI assistant sessions

    Check Point says the Windows malware also steals browser, VPN, wallet, and password-manager data from one endpoint.

  6. · 2d ago GBHackers

    Fake ChatGPT and Claude GitHub apps sideload Yoba malware

    Signed vendor Setup.exe files load YobaDropper, which plants a SYSTEM scheduled task; an observed payload was XMRig 6.26.0 over Tor.

  7. · 2d ago CSO Online

    Fake ChatGPT and Claude apps spread malware via signed binaries

    Check Point’s Yoba loader fetched XMRig and can also retrieve other Windows executables or PowerShell.

  8. · 2d ago Infosecurity Magazine

    Fake ChatGPT and Claude apps on GitHub spread Yoba malware

    Check Point says the two-month campaign sideloads a dropper through signed binaries; the payload seen so far is an XMRig miner.

  9. · 1d ago The Hacker News

    Fake ChatGPT and Claude apps sideload miners via YobaLoader

    Check Point says ZIP files pair signed OpenAI or Anthropic Setup.exe files with a malicious winmm.dll that runs XMRig.