Case hugging-face-gpu-mining-returnsAI supply chainWorldwide
Hugging Face model malware campaign returns to hijack GPUs for crypto mining
Pluto Security found 31 malicious Hugging Face model repositories across seven accounts, published in two waves since September, that abuse the trust_remote_code option to install crypto miners and backdoors on GPU servers. The campaign returned within days of the first takedown, and a near-identical copy uploaded on 25 October was still live when BleepingComputer checked. The most popular model was downloaded over 28,000 times. BleepingComputer also reports, as a separate campaign, a fake Solana starter kit on GitHub aimed at tricking AI coding agents.
HighAttack on AIStatus: OngoingVerified case file
Open in the interactive mapCite this case
- Reported
- Oct 29, 2026
- Location
- Worldwide
- Target
- Hugging Face model users and GPU servers
- Severity
- High
- Verification
- Reported by credible press or researchers
- Triggered by
- Unknown
- Actor type
- Unknown
- Sector
- Technology & AI vendors
- Reported loss
- None reported
- CVE
- None
What happened
Pluto Security found 31 malicious Hugging Face model repositories across seven accounts, published in two waves since September, that abuse the trust_remote_code option to install crypto miners and backdoors on GPU servers. The campaign returned within days of the first takedown, and a near-identical copy uploaded on 25 October was still live when BleepingComputer checked. The most popular model was downloaded over 28,000 times. BleepingComputer also reports, as a separate campaign, a fake Solana starter kit on GitHub aimed at tricking AI coding agents.
Impact: GPU servers were hijacked and backdoors planted; one malicious model was downloaded more than 28,000 times. Total number of victims is Unknown.
Who triggered it
Actor: Unknown
Attribution: Unknown; the summary names no actor.
Motive: Crypto mining using hijacked GPU compute; a separate fake kit aims to install a NetSupport RAT.
How it happened
Attack vector: Malicious models hosted on Hugging Face that compromise GPU servers; a fake Solana starter kit on GitHub aimed at AI coding agents.
Root cause: Unknown; the summary points to malicious models being hosted and downloaded from a public AI model hub.
Timeline
- Reported by BleepingComputer
Response
The campaign returned days after an earlier takedown, so the takedown did not stop it. No other response is stated.
Current status: Ongoing
Research warnings this case matches
- Seen in real incidents AI systems themselves become targets · International AI Safety Report
- Seen in real incidents Privacy and security · MIT AI Risk Repository
Compared with published AI risk research from several institutes on the Research vs reality page.
Sources (open in a new tab)
- BleepingComputer bleepingcomputer.com
- BleepingComputer bleepingcomputer.com
- Runtime Wire runtimewire.com
Cite this case
Hugging Face model malware campaign returns to hijack GPUs for crypto mining. AI Threat Watch, case hugging-face-gpu-mining-returns, Oct 29, 2026. https://aithreat.io/cases/hugging-face-gpu-mining-returns/
Permanent link: https://aithreat.io/cases/hugging-face-gpu-mining-returns/
Related cases
The same type of incident, most recent first.