Reported by 7 outlets

Timeline

1. original reporting

   PoeLLM malware infects exposed AI servers in cryptomining attacks

   bleepingcomputer.com · 6d ago · October 7th, 2026

   A cryptomining campaign targeting exposed AI services is using PoeLLM malware to turn compromised servers into scanners and exploit launchpads. The malware features an uncommon method to retrieve command-and-control (C2) addresses by extracting keywords in a poem hosted on GitHub. Researchers at Lumen's Black Lotus Labs (BLL) tracking the botnet malware say it has compromised more than 3,400 servers, with peak activity reaching as many as 800 infected systems active on a single day. PoeLLM has been active since at least April, but its activity has increased significantly since then, with at least 11 C2 servers spun up to date. According to the research, the operation targeted systems across the United States and Western Europe. Many of those victims run exposed AI tools such as LiteLLM and Ollama, the Gotenberg PDF converter, and the Gitea development toolkit, while signs of Ivanti Sentry targeting were also uncovered. BLL notes that AI/LLM implementations are attractive targets for threat actors because they are often poorly configured, exposed online, and typically run on powerful GPU clusters that are suitable for cryptomining. In a report today , BLL researchers say that PoeLLM, an ELF file named libgcrypt, retrieves four words or phrases from a poem titled “On the Nature of Connection” in a ‘dash.css’ file hosted in a GitHub repository that appears to fork Node.js. It then maps these words to numbers using a hard-coded dictionary, generating a IPv4 address corresponding

2. update from cyberscoop.com

   PoeLLM malware has assembled a sweeping botnet, taking technical cues from a poem

   cyberscoop.com · 6d ago · October 7th, 2026

   Malware that takes technical cues from a poem to assemble a growing botnet by targeting open-source AI services has compromised more than 3,400 servers since April, Lumen Technologies’ Black Lotus Labs said in a report Wednesday. The poem, which a threat actor wrote and posted on a GitHub repository, seems innocuous but it’s driving a stealthy piece of malware researchers call PoeLLM. Four specific words extracted from the poem, which have been changed at least a dozen times, are converted into a command-and-control server address through a hard-coded dictionary embedded in the malware — a framework that bolsters PoeLLM’s resiliency. “The most interesting piece of the poem approach is that the IP address used for C2 communications is invisible outside of the victim’s netflow. In other malware samples, there might be a hard-coded fallback C2 address, or URL which points to an IP address written elsewhere on the internet,” Ryan English, information security engineer at Black Lotus Labs, told CyberScoop. “To anyone who comes across it, this is simply a poem on GitHub. It has no links, no files to download, no encrypted text that could easily be flagged as malicious, even by advanced models,” he added. “There would be no reason for any security researcher to identify this poem as malicious — or know about the IP address hidden within it — unless they had access to the malware referencing it.” Researchers first encountered PoeLLM infrastructure in June during an investigation into

3. update from www.bleepingcomputer.com

   🤖 PoeLLM: cryptomining botnet targeting exposed AI servers (2,100+ infected, ~800 active/day). C2 addresses are hidden in keywords of a poem

   www.bleepingcomputer.com · 6d ago · October 7th, 2026

   C2 addresses are hidden in keywords of a poem

   via Mastodon https://www.bleepingcomputer.com/news/security/poellm-malware-infects-exposed-ai-servers-in-cryptomining-attacks/

4. update from fawkes.rocks

   PoeLLM malware hijacks exposed AI servers for cryptomining https:// fawkes.rocks/2026/10/07/poellm -malware-hijacks-exposed-ai-servers-for-c

   fawkes.rocks · 6d ago · October 7th, 2026

   PoeLLM malware hijacks exposed AI servers for cryptomining https:// fawkes.rocks/2026/10/07/poellm -malware-hijacks-exposed-ai-servers-for-c

   via Mastodon https://fawkes.rocks/2026/10/07/poellm-malware-hijacks-exposed-ai-servers-for-cryptomining/

5. update from www.theregister.com

   Patch Ivanti Sentry against CVE-2026-10520 and scan your environment for PoeLLM infections before the next stanza drops. Reward: You've rece

   www.theregister.com · 6d ago · October 7th, 2026

   Patch Ivanti Sentry against CVE-2026-10520 and scan your environment for PoeLLM infections before the next stanza drops. Reward: You've rece

   via Mastodon https://www.theregister.com/security/2026/10/07/poetry-is-the-new-ai-security-threat-as-poellm-malware-infects-3k-servers/5301672

6. mentioned by www.theregister.com

   The Register: Poetry is the new AI security threat as PoeLLM malware infects 3K+ servers "...A suspected Italian attacker armed with a malwa

   www.theregister.com · 6d ago · October 7th, 2026

7. update from gbhackers.com

   PoeLLM Malware Hijacks 3,400+ Servers for Crypto Mining and Botnet Expansion

   gbhackers.com · 5d ago · October 8th, 2026

   Active since April 2026, the operation combines vulnerability exploitation, cryptocurrency miners and an unusual command-and-control mechanism that derives server addresses from a poem hosted on GitHub. Victims predominantly run LiteLLM, Ollama, Gotenberg and Gitea, […] The post PoeLLM Malware Hijacks 3,400+ Servers for Crypto Mining and Botnet Expansion appeared first on GBHackers Security | #1 Globally Trusted Cyber Security News Platform .

8. update from www.ictjournal.ch

   PoeLLM malware hides its command server in a poem and has compromised more than 3,400 servers

   www.ictjournal.ch · 5d ago · October 8th, 2026

   According to Black Lotus Labs, Lumen's security research team, PoeLLM is a cryptojacking and self-propagating malware that exploits vulnerable AI services exposed to the internet, primarily servers running LiteLLM and Ollama. PoeLLM stands out for how it resolves its C2: it extracts four words or phrases from a poem titled "On the Nature of Connection" published in a GitHub repository, converts them to numbers using a dictionary built into its code, and the four values form the C2 server's IPv4 address. The operator can migrate infrastructure simply by editing words in the text; the poem has been reworked eleven times since first publication, each version steering infected machines to a new command server. Black Lotus Labs told The Register they had never before observed this kind of "adversarial poetry" in a real attack: to anyone who stumbles on it, it is just a poem on GitHub. The technique differs from a classic LLM jailbreak in that no model interprets the poem; the malware uses it directly to recover its C2 address. via Mastodon https://www.ictjournal.ch/news/2026-10-08/un-malware-retrouve-son-serveur-de-commande-dans-un-poeme-et-compromet-plus-de-3400

9. mentioned by www.bleepingcomputer.com

   # PoeLLM # malware infects exposed # AI servers in # cryptomining attacks https://www. bleepingcomputer.com/news/secu rity/poellm-malware-in

   www.bleepingcomputer.com · 5d ago · October 8th, 2026

10. update from socfortress.medium.com

    Cyber Threat Assessment: PoeLLM Botnet Operations and AI Infrastructure Risks

    socfortress.medium.com · 5d ago · October 8th, 2026

    Intro A novel malware campaign known as PoeLLM has successfully compromised thousands of servers by hiding its command-and-control infrastructure within an innocent-looking GitHub poem . Security researchers at Black Lotus Labs discovered that the malicious software extracts specific words from the verse to dynamically calculate network addresses, bypassing conventional security detection tools. This sophisticated botnet primarily targets open-source artificial intelligence services for illicit activities such as exploit scanning and cryptocurrency mining. Furthermore, investigators noted that the operation establishes a resilient network of AI-enabled proxies that continually rotate their connections without requiring code updates. While the full scope of the threat actor’s objectives remains under investigation, linguistic clues in the code suggest possible Italian origins . Threat Overview The rapid operational adoption of open-source artificial intelligence (AI) framework utilities and proxy management software has introduced a novel vector for threat actors seeking to expand malicious botnet infrastructure. The PoeLLM malware campaign represents a critical inflection point for cybersecurity defenders, illustrating how threat actors can leverage non-traditional command-and-control (C2) mechanisms combined with aggressive targeting of exposed AI services. By concealing operational network coordinates within benign public text on GitHub and systematically compromising open-

11. mentioned by fawkes.rocks

    PoeLLM botnet hides C2 address in a GitHub poem https:// fawkes.rocks/2026/10/09/poellm -botnet-hides-c2-address-in-a-github-poem/

    fawkes.rocks · 5d ago · October 8th, 2026