CVE-2026-122014
Deferred Deferred - Pending Action

Code Injection in NPC-Worldwide npcpy

Vulnerability report for CVE-2026-122014, including description, CVSS score, EPSS score, affected products, exploitability, helpful resources, and attack-flow context.

Publication date: 2026-11-04

Last updated on: 2026-11-04

Assigner: VulDB

Description

A vulnerability has been found in NPC-Worldwide npcpy up to 2.1.27. This affects the function load_sft_model of the file npcpy/ft/sft.py. The manipulation of the argument base_model_name_or_path leads to code injection. The attack can be initiated remotely. The exploit has been disclosed to the public and may be used.

CVSS Scores

EPSS Scores

Probability:
Percentile:

Meta Information

Published
2026-11-04
Last Modified
2026-11-04
Generated
2026-11-05
AI Q&A
2026-11-05
EPSS Evaluated
N/A
NVD
EUVD

Helpful Resources

Exploitability

CWE
CWE Icon
KEV
KEV Icon
CWE ID Description
CWE-74 The product constructs all or part of a command, data structure, or record using externally-influenced input from an upstream component, but it does not neutralize or incorrectly neutralizes special elements that could modify how it is parsed or interpreted when it is sent to a downstream component.
CWE-94 The product constructs all or part of a code segment using externally-influenced input from an upstream component, but it does not neutralize or incorrectly neutralizes special elements that could modify the syntax or behavior of the intended code segment.

Attack-Flow Graph

AI Quick Actions

Instant insights powered by AI
Executive Summary

This vulnerability is a code injection flaw in the npcpy library up to version 2.1.27. It occurs in the load_sft_model function of npcpy/ft/sft.py where the base_model_name_or_path argument is manipulated to execute arbitrary code. The attack is remote and exploits hardcoded trust_remote_code=True, allowing malicious Python files to run during model loading.

Detection Guidance

Check for npcpy installations with versions up to 2.1.27. Inspect Python environments for the presence of npcpy/ft/sft.py and review any adapter configurations loaded from untrusted sources. Look for files like model.safetensors or adapter_config.json in directories where models are stored.

Impact Analysis

An attacker could remotely execute arbitrary code on your system by tricking you into loading a malicious adapter from sources like Hugging Face Hub. This could lead to full system compromise, data theft, or further network infiltration if npcpy is used in production environments.

Compliance Impact

This vulnerability could lead to unauthorized code execution, potentially violating GDPR's integrity principle or HIPAA's security rule by exposing sensitive data. Organizations using affected npcpy versions may face compliance breaches if attackers access personal or health data through this exploit.

Mitigation Strategies

Update npcpy to the latest version where trust_remote_code is disabled by default. Avoid loading adapters from untrusted sources. Validate base model paths and disable dynamic imports. Monitor for suspicious activity in model loading processes.

Chat Assistant

Ask questions about this CVE
Hi! I’m here to help you understand CVE-2026-122014. Ask me anything about the vulnerability, its impact, or mitigation strategies.
0/70

EPSS Chart

Security Radar

News mentioning this vulnerability
Loading related security news...