PoeLLM cryptomining malware hid C2 addresses in a GitHub poem, hitting 3,400+ servers
Lumen's Black Lotus Labs detailed a campaign it calls Canto Incognito, in which cryptomining malware named PoeLLM has infected more than 3,400 servers since at least April. Many victims appear to be internet-exposed open-source AI/LLM services such as LiteLLM and Ollama, along with the PDF-conversion API Gotenberg and self-hosted Git platform Gitea; the enterprise gateway Ivanti Sentry may also have been targeted. The malware decoded command-and-control addresses from four words in a two-stanza GitHub poem titled "On the Nature of Connection," revised 11 times so far, and dropped XMRig-based miners that turn infected hosts into scanners and exploit servers. Researchers stressed the poem is only C2 obfuscation, not prompt injection or a jailbreak style of "adversarial poetry."
Exposed AI services can offer attackers both valuable data and GPU access, making AI infrastructure an increasingly attractive target.
美國電信業者Lumen旗下資安研究團隊Black Lotus Labs揭露PoeLLM惡意程式攻擊行動，至少從今年4月起鎖定對外開放的AI及開源服務，累計入侵超過3,400臺伺服器。遭入侵的主機被用於加密貨幣挖礦，也會掃描及攻擊其他系統。
PoeLLM還利用存放在GitHub儲存庫中的英文詩取得指揮與控制（C2）伺服器位址，駭客只要修改詩中的關鍵字，就能讓受害主機連上新的控制伺服器。
受害主機主要分布在美國與西歐，不少運行LiteLLM、Ollama等AI服務，另有PDF轉換工具Gotenberg及軟體開發平臺Gitea。攻擊者會先掃描暴露在網際網路上的服務，再利用安全漏洞，要求目標伺服器從C2下載惡意程式，部分遭入侵的主機隨後被當成攻擊跳板，繼續搜尋及入侵其他伺服器，擴大殭屍網路規模。
PoeLLM的特殊之處，在於利用一首存放在GitHub儲存庫中的英文詩取得C2位址，駭客將該英文詩藏在名為dash.css的檔案中，惡意程式會讀取詩中4個指定位置的字詞，透過程式內建的對照表轉換成4個數字，再組合成C2伺服器的IP位址。當駭客更換控制伺服器，只要修改英文詩中的關鍵字，受害主機就能計算出新的連線位址，不必重新散布惡意程式。自4月13日首次上傳以來，該英文詩已修改11次。
#PoeLLM#Black Lotus Labs#Lumen#LiteLLM#Ollama#Gotenberg#Gitea#Ivanti Sentry#XMRig#Kryptex