CVE-2026-1669Disclosure(keras / keras)

LOWCVSS 7.5 · HIGH

Signal is active with 1 mentions in latest observed window

Immediate actions

Track advisory updates for patch or workaround availability

Recommended action window: Monitor and triage in normal cycle

NVD description

Arbitrary file read in the model loading mechanism (HDF5 integration) in Keras versions 3.0.0 through 3.13.1 on all supported platforms allows a remote attacker to read local files and disclose sensitive information via a crafted .keras model file utilizing HDF5 external dataset references.

keras CVEs Back to dashboard

0.0/ 10 priority

Sources & remediation

Vendor / third-party advisories

https://github.com/google/security-research/security/advisories

Other references

Weakness type (CWE)

CWE-73CWE-200

Priority

LOW

Exploitation

NONE

PoC

NONE

Patch

NONE

Momentum

STABLE

Are you affected?

If you run products in this scope, you should treat this CVE as relevant to your environment.

keras

Threat summary

3 mentions across 3 observed days
Momentum state: stable

What's happening

Technical details provided in 3 signals
Disclosure: 3 classified signals
Peaked 2d ago at 1 mentions (2026-02-11); latest day: 1
3 total mentions across 3 days

Affected systems

Vendors

keras

Products

keras

Deep dive

Activity timeline3 mentions / 3d

Signal classification1 categories

Disclosure

3100.0%

Referenced assets3 URLs

By indicator

Full discourse3 posts

Giuseppe `N3mes1s`@N3mes1s233d
Disclosure
First #Pruva CVE assigned today in Keras. CVE-2026-1669 - Local File Disclosure via HDF5 External Storage During Keras Weight Loading - https://github.com/keras-team/keras/security/advisories/GHSA-3m4q-jmj6-r34q - https://www.cve.org/CVERecord?id=CVE-2026-1669 Not only hashtag#pruva is capable to reproduce, but even doing hyphotesis generation and reproduce them to make sure they are really vulnerabilities or not. And the way to trigger it and read the leaked content is even funnier.
Post summary
The post announces the assignment of CVE-2026-1669 to Keras, linking to the official advisory and CVE record, and notes that the vulnerability allows local file disclosure during weight loading.
000401.0K
12.8K followersView on X
cvereports@_cvereports232d
Disclosure
CVE-2026-1669: Model Poisoning: Turning Keras Weights into Weaponized File Readers A high-severity Arbitrary File Read vulnerability in the Keras machine learning library allows attackers to exfiltrate sensitive local files (like /etc/passwd or AWS cr... https://cvereports.com/reports/CVE-2026-1669
Post summary
The text announces a new high‑severity arbitrary file read vulnerability in Keras (CVE‑2026‑1669) without providing a PoC, exploit code, or patch details.
0000049
26 followersView on X
CVE@CVEnew239d
Disclosure
CVE-2026-1669 Arbitrary file read in the model loading mechanism (HDF5 integration) in Keras versions 3.0.0 through 3.13.1 on all supported platforms allows a remote attacker to read… https://www.cve.org/CVERecord?id=CVE-2026-1669
Post summary
The statement announces CVE‑2026‑1669, describing an arbitrary file read vulnerability in Keras’s HDF5 model loading, but provides no evidence of PoC, exploitation, or patch.
00000194
56.5K followersView on X

CPE platform detail1 entries

Part:

Vendor:

Product:

Version:

Target SW:

Target HW:

1 of 1 entries

Part	Vendor	Product	Version	Target SW	Target HW
App	keras	keras	-	-	-

Explore more

Back to dashboard keras CVEs keras vulnerabilities