Skip to content

Security: VaradaGovind/algorithmic-fairness-analysis

Security

SECURITY.md

Security Policy

Reporting Security Issues

If you discover a security vulnerability, credential leak, or sensitive data exposure risk in this repository, please report it responsibly.

Responsible Disclosure Process

  • Do NOT open a public GitHub issue to report potential security vulnerabilities or sensitive data exposures.
  • Please use GitHub Private Vulnerability Reporting via the repository's Security Advisories tab:
    1. Navigate to the repository on GitHub.
    2. Click on the Security tab.
    3. Under "Reporting", click Report a vulnerability to open a private advisory.

This allows investigation and resolution of the issue before public disclosure.


Scope and Guidelines

In the context of this educational algorithmic fairness framework:

  1. Secret or Credential Leaks: Unintended inclusion of API keys, authentication tokens, or private credentials in code, configuration, or documentation.
  2. Proprietary or Personal Data: Committing or tracking authentic proprietary records or Personally Identifiable Information (PII) within version control.
  3. Evaluation Integrity Safeguards: Flaws in feature firewalls, split isolation, or invariant checks that could lead to silent evaluation leakage or data corruption.
  4. Code Execution Vulnerabilities: Unsafe deserialization or arbitrary code execution via file ingestion or serialization handlers.

Data Notice

  • Never submit proprietary datasets, production logs, or private data in GitHub issues, pull requests, or public discussions.
  • All evaluation workflows in this repository are designed to operate on controlled synthetic data or documented public benchmark datasets (such as the UCI Adult Census dataset).
  • Any local user datasets should remain strictly outside Git version control in accordance with .gitignore.

There aren't any published security advisories