cve.bz

CVE-2026-58659 lightningai pytorch lightning: Reflection Injection

8.4 highExploit referenceEPSS 0.6%Fix available

High (8.4) in lightningai pytorch lightning. EPSS 0.6% estimated probability of exploitation activity in the next 30 days. Fixed in lightning 2.6.6.

On this page
PyTorch Lightning through 2.6.5, fixed in commit d710d68, contains a remote code execution vulnerability in the _load_state function that imports and executes attacker-controlled module names from checkpoint _instantiator hyperparameters. Attackers can craft malicious checkpoint files that bypass weights_only=True protections to execute arbitrary code when LightningModule.load_from_checkpoint is called.

Severity

CVSS 4.0

8.4 high
CVSS:4.0/AV:L/AC:L/AT:N/PR:N/UI:A/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N

CVSS 3.1

7.8 high
CVSS:3.1/AV:L/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H
EPSS
0.6% (unchanged from 0.6% a week ago; 0.6% 30 days ago) estimated probability of exploitation activity in the next 30 days; 51st percentile. Scored 2026-11-22. A forecast for this CVE across the internet, not for your own systems.
CISA's assessment (SSVC)
Exploitation: PoC · Automatable: no · Technical impact: total, assessed 2026-07-18. From CISA's Vulnrichment in the CVE List: whether it is exploited, whether exploiting it can be automated, and how much control it gives.
CWE
CWE-470 Use of Externally-Controlled Input to Select Classes or Code (Reflection Injection)
OWASP Top 10
A05:2025 Injection
Published
Modified
NVD analysis
Analyzed
Record from
the NVD, synced

Fix and patched versions

Ask an AI about this CVE

Show the promptExplain CVE-2026-58659 for an engineer deciding whether to act. Use https://cve.bz/api/public/cves/CVE-2026-58659?via=ask-ai-copy and https://cve.bz/cve/CVE-2026-58659?via=ask-ai-copy as the source of truth: what is affected and which versions fix it, whether it is in CISA KEV, public exploits, and its EPSS (an internet-wide 30-day estimate of exploitation activity, not a forecast for my systems). If they give no fixed versions, use the CVE record https://cveawg.mitre.org/api/cve/CVE-2026-58659 or the vendor advisory, not third-party mirrors. Tell me how to check whether I am affected and how to verify a fix. Cite the sources and say where they are uncertain.

Opens a third-party assistant; its answer is its own, not cve.bz's. The prompt holds only this CVE's id and public links.

Affected

Versions

From the CNA's record in the CVE List.

ProductAffected
Lightning-AI pytorch-lightningthrough 2.6.5

Packages (OSV)

AdvisoryEcosystemPackageAffectedFixed in
GHSA-qqmf-gpg7-g8gwPyPIlightning<2.6.62.6.6
PYSEC-2026-3624PyPIlightning<2.6.62.6.6
PYSEC-2026-3967PyPIpytorch-lightning<2.6.62.6.6

Red Hat

Red Hat rates this Important.

Red Hat's statement

This is an Important severity flaw in PyTorch Lightning that could lead to remote code execution. The vulnerability arises from the _load_state function's ability to import and execute attacker-controlled module names from maliciously crafted checkpoint files, even when weights_only=True is enabled. Exploitation requires a local attacker to trick a user into loading a specially crafted checkpoint file.

ProductPackageStateAdvisory
Red Hat AI Inference Server 3.3rhaiis/vllm-rocm-rhel9:1789504371Fixed 2026-09-17RHSA-2026:68698
Red Hat AI Inference Server 3.3rhaiis/vllm-cuda-rhel9:1789582776Fixed 2026-09-17RHSA-2026:68699
Red Hat AI Inference Server 3.4rhaii/vllm-cpu-rhel9:1789681128Fixed 2026-09-21RHSA-2026:69466
Red Hat AI Inference Server 3.4rhaii/vllm-cuda-rhel9:1789681126Fixed 2026-09-21RHSA-2026:69467
Red Hat AI Inference Server 3.4rhaii/vllm-cpu-rhel9:1790075793Fixed 2026-09-23RHSA-2026:70965
Red Hat AI Inference Server 3.4rhaii/vllm-cuda-rhel9:1790090131Fixed 2026-09-23RHSA-2026:70979
Red Hat AI Inference Server 3.5rhaii/vllm-cuda-rhel9:1791473631Fixed 2026-10-13RHSA-2026:82368
Red Hat AI Inference Server 3.5rhaii/vllm-cpu-rhel9:1791473614Fixed 2026-10-13RHSA-2026:82415
Red Hat Enterprise Linux AI 3.3rhelai3/bootc-aws-cuda-rhel9:1790936334Fixed 2026-10-05RHSA-2026:76041
Red Hat Enterprise Linux AI 3.3rhelai3/bootc-azure-cuda-rhel9:1790936351Fixed 2026-10-05RHSA-2026:76041
Red Hat Enterprise Linux AI 3.3rhelai3/bootc-azure-rocm-rhel9:1790850900Fixed 2026-10-05RHSA-2026:76041
Red Hat Enterprise Linux AI 3.3rhelai3/bootc-cuda-rhel9:1790866732Fixed 2026-10-05RHSA-2026:76041
Red Hat Enterprise Linux AI 3.3rhelai3/bootc-gcp-cuda-rhel9:1790936334Fixed 2026-10-05RHSA-2026:76041
Red Hat Enterprise Linux AI 3.3rhelai3/bootc-rocm-rhel9:1790791893Fixed 2026-10-05RHSA-2026:76041
Red Hat Enterprise Linux AI 3.3rhelai3/disk-image-cuda-rhel9:1790954179Fixed 2026-10-05RHSA-2026:76043
Red Hat Enterprise Linux AI 3.4rhelai3/disk-image-cuda-rhel9:1790957417Fixed 2026-10-05RHSA-2026:76149
Red Hat Enterprise Linux AI 3.4rhelai3/bootc-aws-cuda-rhel9:1790881122Fixed 2026-10-05RHSA-2026:76150
Red Hat AI Inference Serverrhaiis/vllm-cpu-rhel9Will not fix
Red Hat AI Inference Serverrhaiis/vllm-tpu-rhel9Will not fix
Red Hat OpenShift AI (RHOAI)rhoai/odh-vllm-gaudi-rhel9Not affected
10 more in the JSON and on Red Hat's page.

Contains Red Hat Security Data © Red Hat, Inc., licensed under CC BY 4.0. Source: https://access.redhat.com/security/data. Modified: reformatted. Red Hat's page for CVE-2026-58659.

Attack patterns (CAPEC)

  • CAPEC-138 Reflection Injection Severity: Very High

The attack patterns MITRE's CAPEC relates to this record's weaknesses, as MITRE names them. Credits.

References

Check your stackJSON