Information disclosure via mismatched Conv tensor ranks
Published Jul 4, 2026 · Updated Jul 4, 2026
Out-of-bounds read in ONNX 1.21 and earlier allows remote authenticated users to disclose process memory through a crafted Conv model. convPoolShapeInference_opset19 derives kernel_shape from the weight rank but sizes dilations from the input rank, then indexes dilations with the larger bound. Exploitation requires the model to undergo Conv shape inference; the demonstrated read is eight bytes, with no reported write or code execution.
Summary
What happened
Out-of-bounds read in ONNX 1.21 and earlier allows remote authenticated users to disclose process memory through a crafted Conv model. convPoolShapeInference_opset19 derives kernel_shape from the weight rank but sizes dilations from the input rank, then indexes dilations with the larger bound. Exploitation requires the model to undergo Conv shape inference; the demonstrated read is eight bytes, with no reported write or code execution.
The record
- CVE
- CVE-2026-14647
- Published
- Jul 4, 2026
- Updated
- Jul 4, 2026
- Vendor
- ONNX
- Product
- ONNX
- Classifications
- CWE-119, CWE-125
- Attack vector
- network
- Privileges
- authenticated
Timeline
How it unfolded
- Jul 4, 2026CVE publishedPublication date reported by the CVE source.
- Jul 4, 2026Record updatedLatest update available in the CVE record.
Exploitability
Present is not the same as exploitable
Compare your product and version with the public record. A matching version still requires validation against your environment.
Is a vulnerable build present?
Compare these published version ranges with your installed build and any vendor patches.
- Affected versionversion=1.0
- Affected versionversion=1.1
- Affected versionversion=1.10
- Affected versionversion=1.11
- Affected versionversion=1.12
- Affected versionversion=1.13
- Affected versionversion=1.14
- Affected versionversion=1.15
- Affected versionversion=1.16
- Affected versionversion=1.17
- Affected versionversion=1.18
- Affected versionversion=1.19
- Affected versionversion=1.2
- Affected versionversion=1.20
- Affected versionversion=1.21
- Affected versionversion=1.3
- Affected versionversion=1.4
- Affected versionversion=1.5
- Affected versionversion=1.6
- Affected versionversion=1.7
- Affected versionversion=1.8
- Affected versionversion=1.9
What conditions does exploitation require?
What is affected?
Attacks
What attackers are doing with it
Daily unique IPs observed by Shadowserver honeypots for known exploited vulnerabilities (KEVs). Missing observations do not establish an absence of attacks.
Weakness, pattern, technique
Public exploit references
- Crafted Conv model reproducerproof of concept · demonstrated
Labels summarize the accepted research assessment. They do not indicate a test against your environment.
Technologies
Your stack
See the directory against your own environment.
Your stack
Check the software in your environment
Book a demo to see how Hinoki identifies affected software and validates exploitability in your environment.
Book a demo