Hugging Face malware campaign returns to hijack GPUs for crypto mining
A crypto-mining campaign abusing Hugging Face models has returned just days after being taken down, targeting GPU servers.
The popular artificial intelligence platform Hugging Face has faced a resurgence of malicious activity. According to BleepingComputer, a crypto-mining campaign that abuses platform models has returned days after a previous takedown, continuing to target high-performance infrastructure.
During this renewed campaign, attackers have been hijacking GPU servers and planting backdoors. Notably, one of the malicious models involved has been downloaded over 28,000 times, highlighting the widespread risk to developers using open-source resources.
Separately, researchers spotted a fake Solana starter kit on GitHub designed to trick AI coding agents. The malicious repository aims to deceive automated coding tools into inadvertently installing the NetSupport RAT remote access Trojan.
These incidents illustrate how threat actors are rapidly evolving their tactics to target modern AI supply chains and developer ecosystems. For IT professionals globally, verifying the integrity of third-party models and packages is more critical than ever.
As the technology sector continues to grow in regions like Uzbekistan, developers relying on open-source repositories and AI-assisted tools must remain vigilant against sophisticated supply chain attacks and adopt rigorous security practices.
