
Cybersecurity researchers at Palo Alto Networks Unit 42 revealed a vulnerability in Google Cloud’s Vertex AI platform that lets attackers exploit AI agents to access sensitive data and compromise cloud environments. The flaw stems from weaknesses in Vertex AI’s permission model, creating a security blind spot that could expose private artifacts and critical information across organizations using the service. Google has yet to announce a fix.
Vertex AI Vulnerability Exposes Google Cloud Data and Private Artifacts