Improper Limitation of a Pathname to a Restricted Directory ('Path Traversal')
CVE-2026-11816
Summary
Keras versions prior to 3.14.0 are vulnerable to a path traversal issue in the archive extraction utilities located in `keras/src/utils/file_utils.py`. The functions `filter_safe_tarinfos()` and `filter_safe_zipinfos()` validate archive member paths against the process current working directory (CWD) instead of the actual extraction destination. When the process runs with CWD set to `/`, which is common in Docker containers, CI/CD runners, and Jupyter environments, the validation boundary becomes the filesystem root, allowing traversal paths to bypass the security check. Additionally, the zip filter contains a bug that causes an `AttributeError` when a blocked entry is encountered, leading to incomplete extraction. Furthermore, Python 3.11 installations lack the `filter="data"` safety net, leaving them entirely reliant on the flawed CWD-based filter. Exploitation of this vulnerability can result in arbitrary file writes outside the intended extraction directory, enabling attackers to overwrite configuration files, inject malicious code, or corrupt machine learning datasets and pipelines.
- LOW
- NETWORK
- HIGH
- UNCHANGED
- REQUIRED
- NONE
- HIGH
- NONE
CWE-22 - Path Traversal
Path traversal (or directory traversal), is a vulnerability that allows malicious users to traverse the server's root directory, gaining access to arbitrary files and folders such as application code & data, back-end credentials, and sensitive operating system files. In the worst-case scenario, an attacker could potentially execute arbitrary files on the server, resulting in a denial of service attack. Such an exploit may severely impact the integrity, confidentiality, and availability of an application.
References
Advisory Timeline
- Published
