dbcve.org
Deserialization of Untrusted DataWeakness · CWE-502

CVE-2025-12058

MEDIUM · 5.9 CVSS v4.0 Published 2025-10-29
Patch available
A vendor patch is available. No clean upgrade release — apply the published patch.
See remediation →
59/100
Remediation priority · Elevated
Patch available

Official description Straight from the sourceThe vendor's or NVD's own wording, published unedited. Authoritative, but often terse — it says what broke, rarely what to do.

NVD · unedited
The Keras.Model.load_model method, including when executed with the intended security mitigation safe_mode=True, is vulnerable to arbitrary local file loading and Server-Side Request Forgery (SSRF). This vulnerability stems from the way the StringLookup layer is handled during model loading from a specially crafted .keras archive. The constructor for the StringLookup layer accepts a vocabulary argument that can specify a local file path or a remote file path. * Arbitrary Local File Read: An attacker can create a malicious .keras file that embeds a local path in the StringLookup layer's configuration. When the model is loaded, Keras will attempt to read the content of the specified local file and incorporate it into the model state (e.g., retrievable via get_vocabulary()), allowing an attacker to read arbitrary local files on the hosting system. * Server-Side Request Forgery (SSRF): Keras utilizes tf.io.gfile for file operations. Since tf.io.gfile supports remote filesystem handlers (such as GCS and HDFS) and HTTP/HTTPS protocols, the same mechanism can be leveraged to fetch content from arbitrary network endpoints on the server's behalf, resulting in an SSRF condition. The security issue is that the feature allowing external path loading was not properly restricted by the safe_mode=True flag, which was intended to prevent such unintended data access.

Technical summary Written by usOur analysis, written from the advisory, the CVSS vector and the affected-version data. It adds context the advisory leaves out, and never invents facts that are not in the source.

dbcve analysis
How this class of weakness works · CWE-502

The application rebuilds objects from attacker-supplied serialized data, and the act of rebuilding can trigger dangerous code paths. In many runtimes this leads straight to remote code execution. The durable fix is to avoid deserializing untrusted input — or to use a strict, type-limited format with integrity checks.

General guidance for the deserialization of untrusted data class — the official description and references above are authoritative for this specific CVE. Want a bespoke review and a reviewed fix? Ask our team →

CVSS breakdown How the score is builtThe industry scoring standard. It rates how the flaw is reached, what it takes to exploit, and what an attacker gains — the score is derived from those, not the other way round.

From the vector
Attack vector
Adjacent
Complexity
High
Privileges
Low
Authentication
X
User interaction
P
Scope
X

CVSS:4.0/AV:A/AC:H/AT:P/PR:L/UI:P/VC:H/VI:L/VA:L/SC:H/SI:L/SA:L/E:X/CR:X/IR:X/AR:X/MAV:X/MAC:X/MAT:X/MPR:X/MUI:X/MVC:X/MVI:X/MVA:X/MSC:X/MSI:X/MSA:X/S:X/AU:X/R:X/V:X/RE:X/U:X

Remediation Closing itWhat it takes to close this. Where a vendor fix exists we point at it; where none exists we say so plainly, and can build one. Effort estimates are scoped from the advisory, not from your codebase.

dbcve · scoped
Patch available Apply the vendor patch
Vendor patch github.com →
Recommended fix Moderate confidence

Keras version containing PR #21751 fix (check keras-team/keras GitHub releases for the version that includes this security patch)

  1. 1. Identify the current Keras version in use by checking pip show keras or import keras; keras.__version__
  2. 2. Navigate to the GitHub repository for keras-team/keras to locate the fixed version containing PR #21751
  3. 3. Upgrade to the Keras version that includes the security fix from PR #21751 by running: pip install --upgrade keras
  4. 4. Verify the upgrade was successful by checking the Keras version again
  5. 5. Test that Model.load_model() with safe_mode=True now properly blocks external vocabulary paths
  6. 6. If using custom model loading workflows, ensure safe_mode=True is explicitly set when loading models from untrusted sources
Caveat Review Keras release notes for version changes; the fix may alter how StringLookup layer handles vocabulary paths in safe_mode

Generated from the published advisory — verify against the referenced sources before acting.

We can apply and verify the patch for your build — typical engagement from $2,850. Get the patch applied

Scan for this in your stack

Free · runs locally
dbcve dependency scanner

Check whether your project pulls in CVE-2025-12058 — or any other known-vulnerable package — straight from your lock files. Free and open source; it runs locally and uploads nothing.

Get it on GitHub Scan in the browser

References Go to the primary sourcePrimary sources — vendor advisories, patches and trackers. Where our summary and a reference disagree, the reference wins.

Primary sources

Practitioner notes

Contributed

Peer-ranked notes from engineers who’ve handled CVE-2025-12058 in production — separate from our analysis above.

No notes yet

Be the first to add a field note for this CVE — a mitigation you’ve verified, a version caveat, or a link to a working fix. Sign in above to contribute.

What this is

A place for practitioners to share what actually worked: a mitigation you’ve tested, a configuration change, a version- or environment-specific caveat, or a link to a verified patch. The most useful notes rise to the top as peers upvote them, so the signal stays high.

What belongs here
  • Verified mitigations, workarounds, and config changes
  • Version or environment caveats, and links to real fixes
  • No weaponised exploit code, or anything meant to cause harm
  • No spam, self-promotion, credentials, or personal data