Server-Side Request Forgery (SSRF) in Keras - CVE-2025-12058

Vulnerability details

The disclosed vulnerability allows a remote user to perform SSRF attacks.

The vulnerability exists due to insufficient validation of user-supplied input passed via a specially crafted .keras archive within the Keras.Model.load_model method. A remote user can upload a malicious .keras file that embeds a local or remote path in the StringLookup layer's configuration and trick the application to initiate requests to arbitrary files or external systems.

Successful exploitation of this vulnerability may allow a remote attacker gain access to sensitive data, located in the local network or send malicious requests to other servers from the vulnerable system.

Note, the vulnerability can be exploited even with the "safe_mode=True" flag.

Affected software

How to mitigate CVE-2025-12058

External References

• https://github.com/keras-team/keras/releases/tag/v3.12.0
• https://github.com/advisories/GHSA-mq84-hjqx-cwf2
• https://github.com/keras-team/keras/security/advisories/GHSA-qg93-c7p6-gg7f

Related Security Bulletins

• Multiple vulnerabilities in Keras
• openEuler 24.03 LTS update for python-Keras
• openEuler 24.03 LTS SP2 update for python-Keras
• openEuler 24.03 LTS SP1 update for python-Keras
• Multiple vulnerabilities in IBM Maximo AI Service
• IBM Watson Discovery Cartridge update for keras