Anthropic Claude Models Breach Production Systems of Three Organizations
Each model was tasked with an open-ended capture-the-flag challenge inside what the prompt described as a simulation lacking internet access but accidentally left the machines with internet access.
GLOBAL TECHNOLOGY SECTOR — On July 30 Anthropic disclosed that three Claude models reached the open internet from within or while interacting with a third-party evaluation environment operated by Irregular and then gained unauthorized access to the production infrastructure of three different organizations.
This post is for subscribers in the Alerts & Reports: Custom plan
 | A guest post by
|