MindsDB Hit by Critical Code Injection Vulnerability
Zero Day Initiative has disclosed a critical remote code execution vulnerability in MindsDB, allowing authenticated attackers to execute arbitrary code.
Zero Day Initiative (ZDI) has publicly disclosed a critical remote code execution (RCE) vulnerability affecting MindsDB, a popular open-source platform for simplifying database management and AI model deployment. The vulnerability, tracked as ZDI-26-707 and assigned the CVE identifier CVE-2026-92207, carries a CVSS score of 8.8, indicating a high level of severity.
The flaw resides within the OpenBBtable class of MindsDB. It stems from a failure to properly validate user-supplied string input before it is used in the execution of Python code. This oversight allows an authenticated attacker to inject malicious code, which is then executed with the privileges of the service account running MindsDB.
Exploitation requires an attacker to first gain authentication to the affected MindsDB instance. Once authenticated, they can craft specific inputs that trigger the code injection, leading to arbitrary code execution on the underlying system. This could enable attackers to compromise the server, steal sensitive data, or use the compromised system as a pivot point for further network intrusions.
The disclosure timeline indicates that ZDI initially reported the vulnerability to the vendor on November 20, 2025. After follow-ups and a notification of intent to publish as a zero-day advisory on March 9, 2026, ZDI proceeded with the coordinated public release on September 16, 2026. This suggests a significant period where the vulnerability may have been known to the vendor but unpatched.
Given the nature of this vulnerability, which allows for arbitrary code execution, the primary mitigation strategy recommended by ZDI is to restrict access to the MindsDB product. This means limiting who can authenticate to the system and ensuring that only trusted users and applications can interact with it. Network segmentation and strict access controls are crucial.
The vulnerability was discovered and reported by Peter Girnus, Demeng Chen, and Brandon Niemczyk of Trend Zero Day Initiative. Their work highlights the ongoing importance of independent security research in identifying and disclosing critical flaws in widely used software.
While specific details on the affected versions of MindsDB were not immediately available in the advisory, users are strongly advised to consult the official MindsDB security advisories and apply any available patches or updates as soon as possible. The high CVSS score and the nature of the vulnerability underscore the critical need for prompt remediation to prevent potential exploitation.
This disclosure adds to a growing list of vulnerabilities found in platforms that integrate AI and data management, emphasizing the need for robust security practices throughout the software development lifecycle, especially for open-source projects that are often widely adopted.