---
cve: "CVE-2026-44484"
severity: "CRITICAL"
cvss: 9.3
epss: "0.4%"
vendor: "lightning-ai"
kev: false
exploited: false
published: "2026-05-14 15:16:48"
tags: [cve, security, critical]
source: tsecurity.de CVE-Dossier
exported: "2026-11-01T12:59:46+01:00"
---

# CVE-2026-44484

> 9.3 CRITICAL · 🧪 PoC

## Beschreibung

PyTorch Lightning is a deep learning framework to pretrain and finetune AI models. Versions 2.6.2 and 2.6.2 have introduced functionality consistent with a credential harvesting mechanism.

## CNA-Record (Kanon, cvelistV5)

- CNA: **GitHub_M**
- State: PUBLISHED
- Stand: 2026-07-15 00:53:59
- CNA-CVSS: **9.3** (`CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N`)

## CISA-Anreicherung (vulnrichment/ADP)

- CISA-SSVC (Coordinator, v2.0.3):
  - Exploitation: **none**
  - Automatable: **yes**
  - Technical Impact: **total**

## CVSS-Vektor

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

| Metrik | Wert | Bewertung |
|---|---|---|
| AV Angriffsvektor | Netzwerk | bad |
| AC Komplexität | Gering | bad |
| PR Privilegien | Keine | bad |
| UI Interaktion | Keine | bad |

## Schwachstellen-Klasse

- **CWE-829** — Inclusion of Functionality from Untrusted Control Sphere
  The product imports, requires, or includes executable functionality (such as a library) from a source that is outside of the intended control sphere.
- **CWE-506** — Embedded Malicious Code
  The product contains code that appears to be malicious in nature.

## Angriffsmuster (CAPEC)

- [CAPEC-175 — Code Inclusion](https://capec.mitre.org/data/definitions/175.html) _(Severity: Very High)_
- [CAPEC-201 — Serialized Data External Linking](https://capec.mitre.org/data/definitions/201.html) _(Severity: High)_
- [CAPEC-228 — DTD Injection](https://capec.mitre.org/data/definitions/228.html) _(Severity: Medium)_
- [CAPEC-251 — Local Code Inclusion](https://capec.mitre.org/data/definitions/251.html) _(Severity: Medium)_
- [CAPEC-252 — PHP Local File Inclusion](https://capec.mitre.org/data/definitions/252.html) _(Severity: Medium)_
- [CAPEC-253 — Remote Code Inclusion](https://capec.mitre.org/data/definitions/253.html)
- [CAPEC-263 — Force Use of Corrupted Files](https://capec.mitre.org/data/definitions/263.html) _(Severity: Medium)_
- [CAPEC-538 — Open-Source Library Manipulation](https://capec.mitre.org/data/definitions/538.html) _(Severity: High)_

## ATT&CK-Techniken

- [T1055 — Process Injection](https://attack.mitre.org/techniques/T1055/)
- [T1195.001 — Supply Chain Compromise: Software Dependencies and Developme](https://attack.mitre.org/techniques/T1195/001/)

## Referenzen

- <https://github.com/Lightning-AI/pytorch-lightning/security/advisories/GHSA-w37p-236h-pfx3>
- <https://access.redhat.com/security/cve/CVE-2026-44484>
- <https://bugzilla.redhat.com/show_bug.cgi?id=2477476>
- <https://security.access.redhat.com/data/csaf/v2/vex/2026/cve-2026-44484.json>

---
_Exportiert aus dem [tsecurity.de CVE-Dossier](https://tsecurity.de/sicherheitsluecken/cve-2026-44484/) · Datenquellen: EUVD (ENISA), NVD, OSV, CISA KEV, FIRST EPSS, Exploit-DB, BSI WID (CSAF)_