Skip to content

[BUG] Security Vulnerability #358

Description

@anandppatil

Describe the bug
Reporting a security vulnerability in ModelScan through private disclosure, in accordance with the repository's SECURITY.md.

Technical details are intentionally withheld from this public issue and will be provided privately to security@protectai.com.

To Reproduce
Withheld from this public issue. Full reproduction steps, scanner output, a self-contained generator, and a proof-of-concept model file will be provided privately.

Expected behavior
ModelScan should identify unsafe executable content embedded in a supported model format rather than reporting the model as containing no issues.

Further details are included in the private report.
Screenshots
N/A
Environment (please complete the following information):
OS: Windows — local validation environment
ModelScan version: 0.8.8
ML framework version: Keras 3.14.1
Model serialization format: Keras Native (.keras)

Additional context
GitHub handle: anandppatil4383
A private report referencing this issue has been sent to security@protectai.com.
Proof-of-concept material will not be published while coordinated disclosure is in progress.

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    bugSomething isn't working

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions