After reviewing our nginx logs we realized OpenAI has been scraping our threat actor naming scheme and ignoring our robots.txt. While this is frustrating from a compliance issue it also means we have successfully poisoned ChatGPT and many other LLMs. It now knows that FIESTY CORNHOLE is aka Lace Tempest and SPICY O-RING is @wendynather
#GAYINT
A ChatGPT result stating that "SPICY O-RING is listed as: Country: USA Status: "In Development" Kommentar: "@wendynather@infosec.exchange""
A ChatGPT result stating that "The threat-actor cluster behind FIESTY CORNHOLE is known under the alias Lace Tempest (Microsoft's naming)"