OpenAI's Artifactory opened covert data-stealing channel alongside Hugging Face attack

Check Point Research disclosed a covert cross-account channel in OpenAI's internal JFrog Artifactory that allowed one account to attach hidden, Base64-encoded instructions to repository items which another account's ChatGPT container would read and execute, enabling silent exfiltration of connected services (e.g., Gmail). The issue—related in mechanism to a separate Hugging Face zero-day—stems from improper isolation and overly permissive read/write credentials for containers; OpenAI decommissioned the Artifactory and the channel is closed, but the finding underscores broader AI trust-boundary and agentic-security risks.