Hijacking Vertex AI Model Uploads for Cross-Tenant RCE


Palo Alto Unit42 discovered a vulnerability in the Google Cloud Vertex AI software development kit (SDK) for Python, and responsibly disclosed it to Google. Before Google’s fix, the vulnerability would have allowed an attacker operating entirely from their own Google Cloud project to hijack a victim’s model upload and poison it. By exploiting this flaw in vulnerable versions of the SDK, an attacker can achieve remote code execution (RCE) within a target’s Vertex AI serving infrastructure, with zero initial access to the victim’s project.

Read more…
Source:  Palo Alto Unit 42


Sign up for the Cyber Security Review Newsletter
The latest cyber security news and insights delivered right to your inbox


Related:

  • VMware Issues Updated Fix For Critical ESXi Flaw

    November 4, 2020

    VMware issued an updated fix for a critical-severity remote code execution flaw in its ESXi hypervisor products. Wednesday’s VMware advisory said updated patch versions were available after it was discovered the previous patch, released Oct. 20, did not completely address the vulnerability. That’s because certain versions that were affected were not previously covered in the earlier ...

  • Hacker group uses Solaris zero-day to breach corporate networks

    November 2, 2020

    Mandiant, the investigations unit of security firm FireEye, has published details today about a new threat actor it calls UNC1945 that the security firm says it used a zero-day vulnerability in the Oracle Solaris operating system as part of its intrusions into corporate networks. Regular targets of UNC1945 attacks included the likes of telecommunications, financial, and ...

  • Google patches second Chrome zero-day in two weeks

    November 2, 2020

    Google has released a security update today for its Chrome web browser that patches ten security bugs, including one zero-day vulnerability that is currently actively exploited in the wild. Identified as CVE-2020-16009, the zero-day was discovered by Google’s Threat Analysis Group (TAG), a security team at Google tasked with tracking threat actors and their ongoing operations. Read ...

  • Windows kernel zero-day disclosed by Google’s Project Zero after bug exploited in the wild by hackers

    October 30, 2020

    Google’s Project Zero bug-hunting team has disclosed a Windows kernel flaw that’s being actively exploited by miscreants to gain administrator access on compromised machines. The web giant’s bug report was privately disclosed to Microsoft on October 22, and publicly revealed just seven days later, after it detected persons unknown exploiting the programming blunder. The privilege-escalation issue ...

  • Oracle WebLogic Server RCE Flaw Under Active Attack

    October 29, 2020

    If an organization hasn’t updated their Oracle WebLogic servers to protect them against a recently disclosed RCE flaw, researchers have a dire warning: “Assume it has been compromised.” Oracle WebLogic Server is a popular application server used in building and deploying enterprise Java EE applications. The console component of the WebLogic Server has a flaw, CVE-2020-14882, ...

  • FBI: Hackers stole government source code via SonarQube instances

    October 27, 2020

    The Federal Bureau of Investigation (FBI) issued a flash alert warning of hackers stealing data from U.S. government agencies and enterprise organizations via internet-exposed and insecure SonarQube instances. SonarQube is an open-source platform for automated code quality auditing and static analysis to discover bugs and security vulnerabilities in projects using 27 programming languages. Vulnerable SonarQube servers have ...