PoeLLM Malware Campaign Hijacks 3,400+ AI Servers Using Poem-Based Command Infrastructure
Sign in✓ Signed in
A sophisticated malware campaign dubbed PoeLLM has compromised over 3,400 servers worldwide since April 2026, targeting exposed AI and developer infrastructure to power a cryptocurrency-mining operation that disguises command-and-control instructions within GitHub-hosted poems. The threat has infected systems running widely-used open-source tools including LiteLLM, Ollama, Gotenberg, Gitea, and Ivanti Sentry instances across the United States and Western Europe, according to research from Lumen Technologies' Black Lotus Labs.
Continue reading
The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.
Already purchased? Sign in✓ Signed in — this article isn’t included in your current plan.
Unlocking the full article...