NextFin

Major Vulnerabilities Discovered in Google Looker Allowing Cross-Tenant RCE and Data Exfiltration

Summarized by NextFin AI
  • Tenable researchers disclosed critical vulnerabilities in Google Looker, which could allow attackers to gain full system control and access sensitive data across various customer environments.
  • The vulnerabilities, termed "LookOut," include a remote code execution (RCE) chain and a SQL injection flaw targeting Looker’s internal management database, posing severe risks to over 60,000 organizations.
  • Google has patched its managed Cloud Looker services, but self-hosted versions face a "patching gap", leaving organizations vulnerable for months.
  • The incident may accelerate federal requirements for enhanced auditing of third-party data modeling tools, prompting a shift towards "Zero Trust" architectures in business intelligence.

NextFin News - Security researchers at Tenable have disclosed a pair of critical vulnerabilities in Google Looker, a premier business intelligence and data analytics platform, that could allow attackers to gain full system control and access sensitive data across different customer environments. The flaws, collectively referred to as "LookOut," include a remote code execution (RCE) chain and a separate SQL injection vulnerability that targets the platform's internal management database. According to Dark Reading, these vulnerabilities pose a severe threat because Looker serves as the "central nervous system" for corporate data in more than 60,000 organizations globally, including major entities like Coinbase and Walmart.

The most alarming discovery involves an exploit chain that allows for cross-tenant RCE. By leveraging a path traversal vulnerability in Looker’s Git-based project structure, researchers were able to trick the system into executing malicious "Git hooks." Although Looker utilizes JGit—a version of Git that typically does not support hooks—the researchers found a method to force the system to use standard Git via specific POST parameters. By triggering a race condition to bypass security overwrites, they successfully executed arbitrary code on the server. In a cloud environment, this level of access is catastrophic; researchers found that shared infrastructure on the Google Cloud Platform (GCP) could allow an attacker to jump from one compromised tenant to access the service account credentials and data of other organizations.

A second vulnerability, tracked as CVE-2025-12743, involves the exfiltration of Looker’s internal database. This database contains sensitive user credentials, configurations, and secrets that are intended to be hidden from users. According to Help Net Security, researchers identified the internal database connection name through system logs and used error-based SQL injection to bypass access controls. By intentionally triggering specific error messages that contained fragments of the database's contents, an attacker could systematically dump the entire internal management repository, providing a roadmap for further lateral movement within a victim's network.

The discovery of these vulnerabilities underscores a growing architectural risk in the business intelligence (BI) sector. As platforms like Looker become more integrated with diverse data sources—often connecting to 20 to 50 different databases ranging from BigQuery to MySQL—they become single points of failure. The ability to execute code on a BI server is not merely an application-level breach; it is a gateway to the entire data estate of an enterprise. The cross-tenant implications are particularly concerning for the SaaS industry, as they challenge the fundamental security assumption of logical isolation in multi-tenant cloud environments.

While Google responded rapidly by patching its managed Cloud Looker services, the burden of security now shifts to organizations running self-hosted or on-premises versions. According to Tenable researcher Liv Matan, patching these systems is often delayed by technical debt, rigid change management windows, and the fear of disrupting critical business dashboards. This "patching gap" creates a window of opportunity for threat actors to exploit known vulnerabilities in organizations that lack a comprehensive asset inventory or robust DevOps pipelines. Data from industry analysts suggests that self-hosted deployments often lag behind cloud-native versions in security updates by an average of three to six months.

Looking forward, the "LookOut" vulnerabilities will likely prompt a re-evaluation of how BI tools handle administrative secrets and internal configurations. We expect to see a shift toward "Zero Trust" architectures for data analytics, where the BI layer is strictly isolated from the underlying infrastructure and service account credentials. Furthermore, U.S. President Trump’s administration has recently emphasized the protection of critical data infrastructure, and this incident may accelerate federal requirements for enhanced auditing of third-party data modeling tools used by government contractors and essential services.

For the 60,000 organizations relying on Looker, the immediate priority is manual remediation. Google has released security bulletin GCP-2025-052, urging self-hosted users to update to versions 25.12.30+, 25.10.54+, or higher. Beyond patching, security teams must implement the principle of least privilege, ensuring that Looker service accounts have only the minimum necessary access to external data warehouses. As cyber-espionage groups increasingly target data aggregation points, the security of the "central nervous system" of corporate intelligence will remain a primary battleground in 2026.

Explore more exclusive insights at nextfin.ai.

Insights

What are the critical vulnerabilities discovered in Google Looker?

What is remote code execution (RCE) in the context of Looker vulnerabilities?

How do the Looker vulnerabilities impact organizations using the platform?

What measures did Google take in response to the Looker vulnerabilities?

What are the implications of cross-tenant RCE in multi-tenant cloud environments?

How do the Looker vulnerabilities reflect current trends in business intelligence security?

What challenges do organizations face in patching self-hosted versions of Looker?

What shifts in data analytics security architectures are anticipated following the Looker vulnerabilities?

How does the SQL injection vulnerability in Looker work?

What are the long-term impacts of the LookOut vulnerabilities on data infrastructure security?

How do Looker vulnerabilities compare to similar issues in other business intelligence platforms?

What role does technical debt play in delaying security updates for Looker?

What are the implications of the U.S. government's focus on data infrastructure security?

What recommendations are made for securing Looker service accounts?

What specific user feedback has been reported regarding the Looker vulnerabilities?

How did researchers exploit vulnerabilities in Looker's Git-based project structure?

What recent updates have been made to Looker following the disclosure of vulnerabilities?

What is the principle of least privilege, and how does it relate to Looker?

Search
NextFinNextFin
NextFin.Al
No Noise, only Signal.
Open App