Vulnerabilities/GHSA-PWQ7-2GVJ-VG9V

GHSA-PWQ7-2GVJ-VG9V: keras PyPI

SSVC · Estimated by CveMateTrack*Monitor closely. May escalate on new signals. CISA has no entry yet, so CveMate computed this from CVSS and exploit data. It is replaced when CISA publishes.

ExploitationNone (derived)

AutomatableNo (derived)

Technical impactTotal (derived)

Mission & well-beingHigh (assumed)

Source: CveMate estimate, same decision tree as CISA.CISA publishes the first three inputs. Mission & well-being depends on your environment, so it is assumed at its highest: the decision is a worst case.

Published

Published 11-08-2025Last modified 12-08-2025

On this page

Summary Timeline Description CVSS metrics EPSS trend Known exploited Public exploits Weaknesses Affected systems References

Key information

Published11-08-2025

Modified12-08-2025

SourceOSV · Pypi

CVSS8.6

EPSSNot scored

KEV0

Exploits0

Vulnerability summary

GHSA-PWQ7-2GVJ-VG9V is a high (CVSS 8.6) vulnerability in keras PyPI.

Overall risk

34/100

0100

Medium · default modelHow is this calculated?

SSVC · Estimated by CveMateTrack*Monitor closely. May escalate on new signals. CISA has no entry yet, so CveMate computed this from CVSS and exploit data. It is replaced when CISA publishes.

ExploitationNone (derived)

AutomatableNo (derived)

Technical impactTotal (derived)

Mission & well-beingHigh (assumed)

CVSS score

8.6/10

010

High · v4.0 (OSV · Pypi), selected by version preference

EPSS scoreNot scoredRisk metrics not yet published.

KEVNot listedNot listed by any tracked catalog.

RansomwareNot flaggedNo ransomware use recorded.

Public exploits0No public exploits recorded

Timeline

11-08-2025
PublishedVulnerability first disclosed.
12-08-2025
Last modifiedVulnerability information updated.

Description

Duplicate Advisory: Keras safe mode bypass vulnerability ### Duplicate Advisory This advisory has been withdrawn because it is a duplicate of GHSA-c9rc-mg46-23w3. This link is maintained to preserve external references. ### Original Description A safe mode bypass vulnerability in the `Model.load_model` method in Keras versions 3.0.0 through 3.10.0 allows an attacker to achieve arbitrary code execution by convincing a user to load a specially crafted `.keras` model archive.