Vertex AI Vulnerability Exposes Google Cloud Data and Private Artifacts
Summary: Palo Alto Networks Unit 42 disclosed a blind spot in Google Cloud's Vertex AI platform where the Per-Project, Per-Product Service Agent (P4SA) tied to an agent built with the Agent Development Kit carries excessive permissions by default. Unit 42 said any call to an agent deployed via Agent Engine invokes Google's metadata service and exposes the service agent's credentials, the host GCP project, the agent's identity and its scopes, which the researchers used to move from the agent's execution context into the customer project and gain unrestricted read access to all Cloud Storage bucket data there.
Key Takeaway: The same stolen P4SA credentials also reached restricted, Google-owned Artifact Registry repositories, letting researchers download container images from private repos that form the core of the Vertex AI Reasoning Engine - proprietary code that Unit 42 says gives an attacker a blueprint for finding further vulnerabilities.
Read the original at The Hacker News ↗
- Topics
- Data Breaches
- Related
- TSMC employee data breach linked to cloud provider Cyber Risks · Nov 15, 2026Data Breach at La Poste Subsidiary Maileva Affects 22 Million People in France Cyber Risks · Nov 15, 2026City of Osaka warns of data breach impacting 1 million as Japan reels from dozens of cyber incidents Cyber Risks · Nov 13, 2026