Skip to content
Fencer Intel
high 8.8

CVE-2026-44513 json

Diffusers: `trust_remote_code` bypass via `custom_pipeline` and local custom components

published 6mo ago modified 14d ago also GHSA-98h9-4798-4q5v , PYSEC-2026-40 , EUVD-2026-30334

cvss
8.8
version 3.1, CNA
cisa kev
Not listed
not in CISA's catalog
epss
1.1%
EPSS 1.1% chance of exploitation in 30 days, percentile 65%
public exploits
0
none known

Summary

Diffusers is the a library for pretrained diffusion models. Prior to 0.38.0, a trust_remote_code bypass in DiffusionPipeline.from_pretrained allows arbitrary remote code execution despite the user passing trust_remote_code=False (or omitting it, which is the default). The vulnerability has three variants, all sharing the same root cause — the trust_remote_code gate was implemented inside DiffusionPipeline.download() rather than at the actual dynamic-module load site, so any code path that bypassed or short-circuited download() also bypassed the security check. DiffusionPipeline.from_pretrained('repoA', custom_pipeline='attacker/repoB', trust_remote_code=False) — the gate evaluated against repoA's file list rather than repoB's, so repoB's pipeline.py was loaded and executed. DiffusionPipeline.from_pretrained('/local/snapshot', custom_pipeline='attacker/repoB', trust_remote_code=False) — the local-path branch never invoked download(), so the gate was never reached and remote code from repoB executed. DiffusionPipeline.from_pretrained('/local/snapshot', trust_remote_code=False) where the snapshot contains custom component files (e.g. unet/my_unet_model.py) referenced from model_index.json — same root cause; the local path skipped download() and custom component code executed. This vulnerability is fixed in 0.38.0.

Patch / fix

  • pypi diffusers 0.38.0
  • Red Hat Red Hat OpenShift AI 3.4 1786611435, 1786611803, 1787076481, 1787077779

Is CVE-2026-44513 in your dependencies? Fencer shows which of your repositories are affected as soon as an advisory lands, and opens the fix as a pull request.

Check with Fencer

Vendor advisories

What the vendors themselves say about their products: affected, fixed or not affected, and what to do.

red hat product security vex CVE-2026-44513 updated cvss 8.8

Diffusers: Diffusers: Arbitrary remote code execution via `trust_remote_code` bypass

  • Red Hat AI Inference Server rhaiis/vllm-tpu-rhel9
    affected: workaround, no fix planned “Will not fix”
  • Red Hat OpenShift AI (RHOAI) rhoai/odh-openvino-model-server-rhel9
    affected: workaround, no fix available
  • Red Hat AI Inference Server 3.4 rhaii/vllm-cpu-rhel9 1789681128, rhaii/vllm-cpu-rhel9 1790075793, rhaii/vllm-cuda-rhel9 1789681126, rhaii/vllm-cuda-rhel9 1790090131
    fixed: vendor fix RHSA-2026:69466
  • Red Hat Enterprise Linux AI 3.4 rhelai3/bootc-cuda-rhel9 1790868254, rhelai3/bootc-aws-cuda-rhel9 1790881122, rhelai3/bootc-gcp-cuda-rhel9 1790881123, rhelai3/disk-image-cuda-rhel9 1790957417, rhelai3/bootc-azure-cuda-rhel9 1790881120
    fixed: vendor fix RHSA-2026:76150
  • Red Hat OpenShift AI 3.4 rhoai/odh-th06-rocm64-torch291-py312-rhel9 1787076481, rhoai/odh-th06-cuda130-torch210-py312-rhel9 1787077779, rhoai/odh-training-rocm64-torch29-py312-rhel9 1786611435, rhoai/odh-training-cuda128-torch29-py312-rhel9 1786611803
    fixed: vendor fix RHSA-2026:60520
  • Red Hat AI Inference Server rhaiis/vllm-rocm-rhel9 Red Hat Enterprise Linux AI 3.4 rhelai3/bootc-rocm-rhel9 1790614447, rhelai3/bootc-azure-rocm-rhel9 1790707552
    not affected: vulnerable code not present
  • Red Hat OpenShift AI (RHOAI) rhoai/odh-th06-cuda130-torch291-py312-rhel9
    not affected: component not present
  • Red Hat OpenShift AI 3.4 rhoai/odh-cli-rhel9 1787330837, rhoai/rhai-cli-rhel9 1787330837, rhoai/odh-automl-rhel9 1786612415, rhoai/odh-mlflow-rhel9 1787226790, rhoai/odh-autorag-rhel9 1786612637, rhoai/odh-trainer-rhel9 1787089049, +96 more
    not affected: vulnerable code not present

Red Hat CSAF VEX © Red Hat, Inc., from https://security.access.redhat.com/data/csaf/v2/vex/; changed: reduced to product statuses and remediations. CC BY 4.0 · csaf

CISA SSVC

Exploitation
none
Automatable
no
Technical impact
total
Assessed
5mo ago

Exploit prediction

1.07% chance of exploitation in the next 30 days, higher than 65% of all scored CVEs (FIRST EPSS, 2026-11-09).

Exploits and checks

  • github.com https://github.com/huggingface/diffusers/security/advisories/GHSA-98h9-4798-4q5v referenced PoC, tagged "exploit" by a source, unverified

CVSS scores

assigned by version score vector
CNA 3.1 8.8 CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H
CISA ADP 3.1 8.8 CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H
GitHub 3.1 8.8 CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H
OSV 3.1 8.8 CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H

Affected packages

ecosystem package affected fixed in
pypi diffusers
<0.38.0
0.38.0

Affected products

  • huggingface diffusers
  • Red Hat Red Hat OpenShift AI 3.4
  • Red Hat Red Hat AI Inference Server
  • Red Hat Red Hat Enterprise Linux AI (RHEL AI) 3
  • Red Hat Red Hat OpenShift AI (RHOAI)

Classification

Weakness
  • CWE-94 Improper Control of Generation of Code
  • CWE-358 Improperly Implemented Security Check for Standard

References

Sources

Fencer Intel keeps one entry per vulnerability: the CVE id wins, then GHSA, and records that alias each other are merged, except that two CVEs always keep separate entries. The same data is available as JSON.