CVE-2024-27132: 
NixOS vulnerability analysis and mitigation

CVE-2024-27132 is a security vulnerability discovered in MLflow, affecting versions up to and including 2.9.2. The vulnerability was identified as an insufficient sanitization issue that leads to Cross-Site Scripting (XSS) when running an untrusted recipe in MLflow. The issue was discovered by JFrog Security Research Team and was publicly disclosed on February 23, 2024 (JFrog Advisory).

The vulnerability stems from lack of sanitization over template variables in MLflow. When running an untrusted recipe in a Jupyter Notebook environment, this vulnerability can lead to client-side Remote Code Execution (RCE). The issue has been assigned a CVSS v3.1 base score of 7.5 (High) by JFrog and 9.6 (Critical) by NIST, with the vector string CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:C/C:H/I:H/A:H (NVD).

The vulnerability can allow attackers to achieve client-side Remote Code Execution (RCE) when running untrusted recipes in Jupyter Notebook. This could potentially enable attackers to perform extensive lateral movement within an organization, as ML clients typically have access to important ML services such as ML Model Registries or MLOps Pipelines (Hacker News).

The vulnerability has been patched in MLflow through a fix that was merged via GitHub pull request #10873. Organizations should upgrade to a version newer than 2.9.2 to protect against this vulnerability (GitHub PR).

The security community has highlighted this vulnerability as part of a broader collection of 22 security shortcomings in open-source machine learning tools and frameworks. Security researchers emphasize the importance of treating all XSS vulnerabilities in ML libraries as potential arbitrary code execution risks, particularly when used with Jupyter Notebook (Hacker News).