CVE-2025-12409

Vulnerability

Description

A SQL injection vulnerability was discovered in Looker Studio, affecting its handling of native functions in calculated fields [1]. The root cause was insufficient filtering of SQL keywords, allowing an attacker to bypass Looker Studio's word filtering by using comments (e.g., /**/SELECT) to inject arbitrary SQL code [2]. This flaw could be exploited when a report was configured with 'Viewer's Viewer's credentials and native functions enabled.

Exploitation

Method

An attacker could create a malicious Looker Studio report containing a crafted calculated field with the data source [2]. The injected SQL query would then be executed in the context of the victim's BigQuery permissions when the victim accessed the report. The attacker could leverage publicly accessible BigQuery tables and GCP logs to perform a blind, character-by-character data exfiltration across tenants [2]. A single click on the malicious report or a website embedding it was sufficient to trigger the exploit.

Impact

Successful exploitation allowed an attacker to exfiltrate any BigQuery data accessible to the victim [1][2]. This posed a significant risk to data confidentiality, as no additional authentication beyond the victim's existing session was required.

Mitigation

Google patched this vulnerability on July 7, 2025 [1]. No customer action was needed as the fix was applied server-side. The vulnerability was disclosed through coordinated responsible disclosure by Tenable Research [2].