---
id: CVE-2026-31224
aliases:
  - GHSA-gpx5-7xm4-229w
  - PYSEC-2026-3070
title: Snorkel MultitaskClassifier.load uses an unsafe torch.load
summary: Snorkel MultitaskClassifier.load uses an unsafe torch.load
severity: high
cvss: 8.8
cvssVector: 'CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H'
vendor: snorkel
product: snorkel
ecosystem: pip
affected:
  - snorkel <= 0.10.0
published: '2026-05-12'
updated: '2026-07-13'
source: OSV
sourceUrl: 'https://osv.dev/vulnerability/GHSA-gpx5-7xm4-229w'
references:
  - url: 'https://nvd.nist.gov/vuln/detail/CVE-2026-31224'
  - url: 'https://github.com/snorkel-team/snorkel'
  - url: 'https://www.notion.so/CVE-2026-31224-35d1e1393188814185f3f6db86c9a4e9'
tags:
  - osv
  - pip
epss: 0.00687
epssPercentile: 0.50631
ingestedAt: '2026-07-13T18:57:58.528Z'
---

## Overview

The snorkel library thru v0.10.0 contains an insecure deserialization vulnerability (CWE-502) in the MultitaskClassifier.load() method of the MultitaskClassifier class. The method loads model weight files using torch.load() without enabling the security-restrictive weights_only=True parameter. This default behavior allows the deserialization of arbitrary Python objects via the Pickle module. A remote attacker can exploit this by providing a maliciously crafted model file, leading to arbitrary code execution on the victim's system when the file is loaded via the vulnerable method.

## Affected packages

- `snorkel <= 0.10.0`

## Remediation

Refer to the advisory for the patched release.