GCD AI RADAR
incidents/AIR-2026-0872

Hugging Face Transformers save_pretrained allows arbitrary file writes

Attack on AI2 Disclosed vulnerabilityInfrastructure vulnerabilitySupply chain

A vulnerability in Hugging Face Transformers permits arbitrary file writes via path traversal in chat template names during model saving. Malicious repositories can exploit tokenizer configurations to write files outside intended directories.

disclosed CVENo confirmed exploitation, so this record is tracked as a disclosed vulnerability and not counted as an incident. It moves into the counts if CISA or the vendor confirms exploitation. Why

Details

type
Attack on AI
An AI model, agent or AI-powered product was the thing exploited, abused or leaked.
categories
  • Infrastructure vulnerabilityprimary
    CVEs and advisories in AI frameworks, inference servers, MCP servers, AI developer tools and hosting stacks.
  • Supply chain
    Malicious or backdoored models, poisoned packages, unsafe deserialization and other compromised AI dependencies.
owasp llm top 10
mitre atlas
Both rows are derived from the type and categories (OWASP 2025 edition, ATLAS release 2026.09), not classified per record. How
severity
2 · Disclosed vulnerability
CVE or vendor advisory, patched or patchable, no confirmed exploitation.
targets
Framework / tooling
products
Transformers
incident date
2 Aug 2026
first reported
2 Aug 2026
origin
GitHub Advisory Database
Vulnerability record from NVD / the GitHub Advisory Database. This product uses the NVD API but is not endorsed or certified by the NVD.
classifier confidence
0.95 · published at or above 0.80; see the methodology
tags
CVE-2026-9856 · transformers · path-traversal
record
created 2026-10-18 07:00Z · updated 2026-10-18 07:00Z · json

Sources · 1

  1. 2026-08-02
    CVE-2026-9856: Transformers save_pretrained path traversal allows arbitrary file writes through chat template names
    GitHub Advisory Database · github.com/advisories/GHSA-xrqw-3rrv-vx5w

Titles and links belong to their publishers. The summary above is written by GCD AI RADAR in its own words; data is CC BY-SA 4.0.

Related incidents