CVE-2026-1462 | Safe Mode Bypass in keras-team/keras

A vulnerability in the `TFSMLayer` class of the `keras` package, version 3.13.0, allows attacker-controlled TensorFlow SavedModels to be loaded during deserialization of `.keras` models, even when `safe_mode=True`. This bypasses the security guarantees of `safe_mode` and enables arbitrary attacker-controlled code execution during model inference under the victim's privileges. The issue arises due to the unconditional loading of external SavedModels, serialization of attacker-controlled file paths, and the lack of validation in the `from_config()` method.

Published: 2026-04-13 Last update: 2026-04-17 Assigner: [email protected] Source: [email protected]
NVD Status：Awaiting Analysis，CVE State: published
View at NVD, CVE.org, EUVD

CVE-2026-1462 is rated Moderate Risk (45.1/100): CVSS High severity, with low exploitation likelihood (EPSS 0.36%). Review affected assets and schedule remediation.

Risk is dynamic; we continuously reassess and refresh what is shown on this page as upstream context changes.

Exploit prediction scoring system (EPSS) score for CVE-2026-1462

EPSS lead: Daily EPSS estimates relative likelihood of exploitation; percentile ranks this CVE among scored vulnerabilities (higher = more severe relative rank).

# Date Old EPSS score New EPSS score Delta (New - Old)
1 2026-06-15 0.06% 0.36% +0.30%
2 2026-04-14 — 0.06% —

Full EPSS history (2 records total)

Common vulnerability scoring system (CVSS) metrics for CVE-2026-1462

CVSS metrics for this CVE.

Base score Version Severity Vector Exploitability Impact Score source
8.8 3.0 HIGH
CVSS:3.0/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H Click to expand
Attack vector (AV:N)
Could be attacked over the internet or any normal routed network—not just someone sitting at the machine.
Attack complexity (AC:L)
Once they can reach the bug, pulling it off is straightforward—no weird race conditions or rare setup.
Privileges required (PR:N)
No account or special rights needed—anonymous or random user is enough.
User interaction (UI:R)
A real person has to do something—click, install, enable—otherwise it doesn’t land.
Scope (S:U)
Damage stays in the same “trust bubble” as the broken component—no big spill into unrelated systems.
Confidentiality (C:H)
Serious risk that confidential data gets exposed in a big way.
Integrity (I:H)
They could widely tamper with or forge data—trust in the data is badly hurt.
Availability (A:H)
Could take the service down hard or make it unusable for people who depend on it.
 2.8 5.9 [email protected]

Weakness enumeration for CVE-2026-1462

GitHub Security Advisory for CVE-2026-1462

GHSA-4f3f-g24h-fr8m · Severity: high · Ecosystem: pip — Keras has an untrusted deserialization vulnerability

OS Trackers for CVE-2026-1462

vendor priority summary link
debian end-of-life CVE-2026-1462 end-of-life priority: Debian including 1 source packages (keras), 1 status rows across 1 suites (bullseye): open 1. https://security-tracker.debian.org/tracker/CVE-2026-1462
redhat high — https://access.redhat.com/security/cve/CVE-2026-1462
ubuntu medium CVE-2026-1462 medium priority: Ubuntu including 1 source packages (keras), 6 status rows across 6 suites (bionic, focal, jammy, noble, questing, upstream): DNE 3, needs-triage 3. https://ubuntu.com/security/CVE-2026-1462

Affected software / configurations for CVE-2026-1462

Vendor Product Version Raw CPE
No affected products in dataset.

References for CVE-2026-1462

cvelogic Threat Intelligence