Google has addressed a significant vulnerability, named “Bad.Build,” in its Cloud Build service that allowed hackers to tamper with application images and potentially infect users.
Although a fix was already released in June, the researchers who discovered the bug recently provided a detailed breakdown, likening it to threats seen in SolarWinds and other supply chain attacks.
The vulnerability stemmed from default service accounts’ permissions within Cloud Build, enabling attackers to impersonate these accounts and manipulate builds by injecting malicious code, leading to potential consequences such as Denial-of-Service attacks, data theft, and malware distribution.
Orca Security, the firm that reported the bug to Google, warned that attackers could exploit this flaw, allowing them to manipulate images in Google’s Artifact Registry and inject malicious code. Such manipulated applications could affect users, leading to supply chain attacks that cross from the supplier’s environment to their customers’ on-premise or semi-SaaS environments.
While Google implemented a fix, security experts raised concerns that it only partially addresses the issue, leaving organizations vulnerable to elevated rights and supply chain risks. Orca Security’s Roi Nisimi explained that further measures must be put in place to adequately protect against these risks.
Despite Orca Security’s assessment, Google defended its access provision to service accounts, stating it is the “nature of automated systems that run independently.” Both Google and Orca Security advised organizations to review permissions and make necessary adjustments based on their security posture and other relevant factors.
By addressing the “Bad.Build” vulnerability, Google aims to prevent potential tampering of application images and secure the Cloud Build service, safeguarding users from malicious attacks and supply chain risks.
