HighVulnerability

CVE-2025-9906: Keras Model.load_model arbitrary code execution via crafted .keras archive

Source: NVD(opens in a new tab)
Identifier: CVE-2025-9906
Published: September 19, 2025
Record updated: October 9, 2026

Affected: keras < 3.11.0
Fixed in: 3.11.0
released July 29, 2025, 51 days before disclosure
Known exploitation: Not listed in the CISA Known Exploited Vulnerabilities catalog at the last check.
EPSS: 0.2%

Summary

CVE-2025-9906 affects the Keras Model.load_model method, which can execute arbitrary code even when safe_mode=True is set. An attacker supplies a specially crafted .keras archive whose config.json calls keras.config.enable_unsafe_deserialization() to disable safe mode, then includes a Lambda layer carrying pickled Python code. Both must be in the same archive, with the deserialization call placed first.

Mitigation

The source does not state a fix yet. Check the original advisory for updates.

Details

Scores and exploitation

CVSS score: 7.3(high)
Patch available: Yes
EPSS: 0.2%Estimated probability that this CVE is exploited in the next 30 days, from the FIRST Exploit Prediction Scoring System. Read on November 4, 2026.

Weaknesses and techniques

Weakness (CWE): CWE-502
Attack pattern (CAPEC): CAPEC-586

Affected

Packages: keras@< 3.11.0 (fixed: 3.11.0)
Vendors: Google

Classification

Attack type: Other
Attack sophistication: Moderate
AI component targeted: Framework
Impact: Confidentiality,
Integrity,
Availability
Labels: security

How this record was produced

Summaries, mitigations and labels on this page are generated automatically from the source. The linked source is authoritative. The source gives this advisory a description and no title, so the title above is a headline generated from the source text.

Original source: https://nvd.nist.gov/vuln/detail/CVE-2025-9906

First tracked: February 15, 2026 at 08:42 PM ET, from NVD/CVE Database

Classifier: claude-haiku-5-5, prompt v4

Classifier confidence: 95% (self-assessed)

Affected packages in the Exposure Registry

Matched by package name and ecosystem. Each entry shows whether the package delegates to a language model and how many tracked packages depend on it.

kerasPyPINo LLM dependency detected · 6 tracked dependents

Summary

Mitigation

Affected packages in the Exposure Registry

Related items