VA-26-233-01
Vulnerability from csaf_cisa - Published: 2026-08-21 15:41 - Updated: 2026-08-21 15:41Summary
ONNX symlink-following and path-traversal arbitrary file write
Notes
Legal Notice: All information products included in [https://github.com/cisagov/CSAF/tree/develop/csaf_files/IT/white](https://github.com/cisagov/CSAF/tree/develop/csaf_files/IT/white) are provided \"as is\" for informational purposes only. The Department of Homeland Security (DHS) does not provide any warranties of any kind regarding any information contained within. DHS does not endorse any commercial product or service, referenced in this product or otherwise. Further dissemination of this product is governed by the Traffic Light Protocol (TLP) marking in the header. For more information about TLP, see [https://us-cert.cisa.gov/tlp/](https://us-cert.cisa.gov/tlp/).
Countries and Areas Deployed: Worldwide
Critical Infrastructure Sectors: Information Technology
Risk Evaluation: A local attacker with write access to the directory where a victim serializes external data can deterministically pre-plant a symlink that is being followed, causing the victim's write to append to any file the victim can write, e.g. ~/.ssh/authorized_keys, cron files, or application configs.
Recommended Practices: Fixed in 1.21.0.
Company Headquarters Location: United States
7.1 (High)
References
4 references
{
"document": {
"category": "csaf_vex",
"csaf_version": "2.0",
"distribution": {
"tlp": {
"label": "WHITE"
}
},
"lang": "en-US",
"notes": [
{
"category": "legal_disclaimer",
"text": "All information products included in [https://github.com/cisagov/CSAF/tree/develop/csaf_files/IT/white](https://github.com/cisagov/CSAF/tree/develop/csaf_files/IT/white) are provided \\\"as is\\\" for informational purposes only. The Department of Homeland Security (DHS) does not provide any warranties of any kind regarding any information contained within. DHS does not endorse any commercial product or service, referenced in this product or otherwise. Further dissemination of this product is governed by the Traffic Light Protocol (TLP) marking in the header. For more information about TLP, see [https://us-cert.cisa.gov/tlp/](https://us-cert.cisa.gov/tlp/).",
"title": "Legal Notice"
},
{
"category": "other",
"text": "Worldwide",
"title": "Countries and Areas Deployed"
},
{
"category": "other",
"text": "Information Technology",
"title": "Critical Infrastructure Sectors"
},
{
"category": "summary",
"text": "A local attacker with write access to the directory where a victim serializes external data can deterministically pre-plant a symlink that is being followed, causing the victim\u0027s write to append to any file the victim can write, e.g. ~/.ssh/authorized_keys, cron files, or application configs.",
"title": "Risk Evaluation"
},
{
"category": "general",
"text": "Fixed in 1.21.0.",
"title": "Recommended Practices"
},
{
"category": "other",
"text": "United States",
"title": "Company Headquarters Location"
}
],
"publisher": {
"category": "coordinator",
"contact_details": "https://www.cisa.gov/report",
"issuing_authority": "CISA",
"name": "CISA",
"namespace": "https://www.cisa.gov/"
},
"references": [
{
"category": "self",
"summary": "Vulnerability Advisory VA-26-233-01 CSAF",
"url": "https://raw.githubusercontent.com/cisagov/CSAF/develop/csaf_files/IT/white/2026/va-26-233-01.json"
}
],
"title": "ONNX symlink-following and path-traversal arbitrary file write",
"tracking": {
"current_release_date": "2026-08-21T15:41:33Z",
"generator": {
"engine": {
"name": "VINCE-NT",
"version": "1.15.0+build.101"
}
},
"id": "VA-26-233-01",
"initial_release_date": "2026-08-21T15:41:33Z",
"revision_history": [
{
"date": "2026-08-21T15:41:33Z",
"number": "1.0.0",
"summary": "Initial publication"
}
],
"status": "final",
"version": "1.0.0"
}
},
"product_tree": {
"branches": [
{
"branches": [
{
"branches": [
{
"category": "product_version_range",
"name": "\u003c1.21.0",
"product": {
"name": "ONNX ONNX \u003c1.21.0",
"product_id": "CSAFPID-0001"
}
},
{
"category": "product_version",
"name": "1.21.0",
"product": {
"name": "ONNX ONNX 1.21.0",
"product_id": "CSAFPID-0002"
}
}
],
"category": "product_name",
"name": "ONNX"
}
],
"category": "vendor",
"name": "ONNX"
}
]
},
"vulnerabilities": [
{
"acknowledgments": [
{
"names": [
"tsigouris007"
]
},
{
"names": [
"kpatsakis"
]
}
],
"cve": "CVE-2026-49114",
"cwe": {
"id": "CWE-367",
"name": "Time-of-check Time-of-use (TOCTOU) Race Condition"
},
"notes": [
{
"category": "summary",
"text": "In ONNX before 1.21.0, the \u0027save_external_data\u0027 function builds the external-data file path from the model\u0027s external_data location field and opens it for writing without \u0027O_NOFOLLOW/O_EXCL\u0027, after a non-atomic \u0027os.path.isfile()\u0027 check. A local attacker with write access to the directory where a victim serializes external data can deterministically pre-plant a symlink that is being followed, causing the victim\u0027s write to append to any file the victim can write, e.g. ~/.ssh/authorized_keys, cron files, or application configs. Fixed in 1.21.0.",
"title": "Description"
},
{
"category": "details",
"text": "SSVCv2/E:P/A:N/T:P/2026-05-27T16:00:11Z/",
"title": "SSVC"
}
],
"product_status": {
"fixed": [
"CSAFPID-0002"
],
"known_affected": [
"CSAFPID-0001"
]
},
"references": [
{
"category": "external",
"summary": "github.com",
"url": "https://github.com/onnx/onnx/security/advisories/GHSA-q56x-g2fj-4rj6"
},
{
"category": "external",
"summary": "VA-26-233-01 CSAF",
"url": "https://raw.githubusercontent.com/cisagov/CSAF/develop/csaf_files/IT/white/2026/va-26-233-01.json"
},
{
"category": "external",
"summary": "CVE-2026-49114",
"url": "https://www.cve.org/CVERecord?id=CVE-2026-49114"
}
],
"release_date": "2026-03-31T00:00:00Z",
"remediations": [
{
"category": "vendor_fix",
"date": "2026-03-31T00:00:00Z",
"details": "Fixed in 1.21.0.",
"product_ids": [
"CSAFPID-0001"
],
"url": "https://github.com/onnx/onnx/security/advisories/GHSA-q56x-g2fj-4rj6"
},
{
"category": "vendor_fix",
"date": "2026-03-31T00:00:00Z",
"details": "Fixed in 1.21.0.",
"product_ids": [
"CSAFPID-0002"
],
"url": "https://github.com/onnx/onnx/security/advisories/GHSA-q56x-g2fj-4rj6"
}
],
"scores": [
{
"cvss_v3": {
"baseScore": 7.1,
"baseSeverity": "HIGH",
"vectorString": "CVSS:3.1/AV:L/AC:L/PR:N/UI:R/S:U/C:N/I:H/A:H",
"version": "3.1"
},
"products": [
"CSAFPID-0001"
]
}
],
"title": "ONNX symlink-following and path-traversal arbitrary file write"
}
]
}
Loading…
Loading…
Experimental. This forecast is provided for visualization only and may change without notice. Do not use it for operational decisions.
Forecast uses a logistic model when the trend is rising, or an exponential decay model when the trend is falling. Fitted via linearized least squares.
Sightings
| Author | Source | Type | Date | Other |
|---|
Nomenclature
- Seen: The vulnerability was mentioned, discussed, or observed by the user.
- Confirmed: The vulnerability has been validated from an analyst's perspective.
- Published Proof of Concept: A public proof of concept is available for this vulnerability.
- Exploited: The vulnerability was observed as exploited by the user who reported the sighting.
- Patched: The vulnerability was observed as successfully patched by the user who reported the sighting.
- Not exploited: The vulnerability was not observed as exploited by the user who reported the sighting.
- Not confirmed: The user expressed doubt about the validity of the vulnerability.
- Not patched: The vulnerability was not observed as successfully patched by the user who reported the sighting.
Loading…
The MITRE ATT&CK techniques below are AI-generated suggestions, inferred from the description of the
vulnerability by the CIRCL/vulnerability-attack-technique-classification-roberta-base
model, served locally by ML-Gateway.
They have not been verified by an analyst and are provided for guidance only.
The approach is described in our paper Mapping CVEs to MITRE ATT&CK Techniques: A Curated Gold-Set Classifier and the Limits of LLM-Assisted Label Expansion.
The approach is described in our paper Mapping CVEs to MITRE ATT&CK Techniques: A Curated Gold-Set Classifier and the Limits of LLM-Assisted Label Expansion.
Loading…
Loading…