RCE in Google Keras
CVE-2025-1550
The Keras Model.load_model function permits arbitrary code execution, even with safe_mode=True, through a manually constructed, malicious .keras archive. By altering the config.json file within the archive, an attacker can specify arbitrary Python modules and functions, along with their arguments, to be loaded and executed during model loading.
Vulnerability class: RCE (Remote Code Execution)
Published · last modified .
CVSS v3 metric
CVSS v3 base score 9.8 (Critical). Vector: CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H.
EPSS exploit prediction
EPSS 0.026, 85.6th percentile (scored ); unchanged over 30 days. About EPSS.
Affected products
- Google Keras — versions 3.0.0
- Keras
Weakness classification (CWE)
Public exploits and proof-of-concept code
1 public repository with exploit or proof-of-concept code for this CVE:
- ChCh0i/cve-2025-1550 (last pushed 2025-10-24)
Also mentioned in 8 other repositories (fardeen-ahmed/Bug-bounty-Writeups, fkie-cad/nvd-json-data-feeds, gpxlnx/medium-writeup, …). These reference the CVE in a list, notes or a scanner and are not exploit code for it.
References
- github.com/keras-team/keras/pull/20751 (Patch, Issue Tracking)
- towerofhanoi.it/writeups/cve-2025-1550/ (Exploit, Third Party Advisory)
This CVE has public exploit evidence. See the latest CVEs with public exploits.