LLMjacking: Stolen Cloud Credentials Used in New AI Attack

A Sysdig blog post on LLMjacking, the use of stolen cloud credentials to run language models. The companion cites it for people who run models on their victims' bills and resell the access, a human-operated form of kleptoparasitism in the food web. It also appears among the stealing cases in Fig. 5a.

Cited in

the companion paper (Companion §4, Fig. 5)

Used for

Stealing (kleptotrophy); food web; Fig. 5a

Strand

Cybersecurity & AI-enabled attack · AI-enabled attacks (threat intelligence)

Cite

@misc{sysdig2024llmjacking,
  title = {LLMjacking: Stolen Cloud Credentials Used in New AI Attack},
  author = {Alessandro Brucato},
  year = {2024},
  howpublished = {Sysdig blog},
  url = {https://www.sysdig.com/blog/llmjacking-stolen-cloud-credentials-used-in-new-ai-attack},
}