Cyber Risks · The Hacker News

Fake OpenAI Privacy Filter Repo Hits #1 on Hugging Face, Draws 244K Downloads

Summary: A malicious Hugging Face repository named Open-OSS/privacy-filter impersonated OpenAI's openai/privacy-filter model, copying its model card nearly verbatim and shipping a loader.py file that fetched and executed a Rust-based information stealer on Windows machines. Before Hugging Face disabled it, the repository reached the #1 trending position with roughly 244,000 downloads and 667 likes within 18 hours, numbers HiddenLayer suspects were artificially inflated.

Why it matters: The loader pulled its payload from a Base64-encoded URL on the JSON Keeper paste service, letting the operators swap malware without touching the repository, and the final stealer targeted Discord, cryptocurrency wallets and seed phrases, FileZilla configurations, and Chromium- and Gecko-based browsers. HiddenLayer tied the shared api[.]eth-fastscan[.]org infrastructure to six more repositories and to a command-and-control server previously used to deliver ValleyRAT, which it attributes to the Chinese group Silver Fox.

Read the original at The Hacker News ↗

Topics
Malware
Related