AI-Assisted Secure Coding (Appendix C) · Adversarial AI Detection in Inbound Contributions
Detect suspicious dependencies in PRs
Verify that automated PR pipeline gates detect known indicators of LLM-generated or LLM-assisted malicious payload patterns: registry-confusable or typosquatted dependency names, package references that do not resolve to any published version, and dependencies whose creation, first-publication, or maintainer-change timestamps look anomalous relative to the PR.
OWASP AISVS v1.01, requirement AC.13.3
In plain English
Check PRs automatically for signs of AI-assisted malicious payloads, such as look-alike package names, dependencies that do not exist, or packages created or handed to new maintainers suspiciously recently.
What good looks like
- Dependency changes in PRs are checked.
- Typosquats and non-existent packages are flagged.
- Recent creation or ownership changes are flagged.
How teams usually do it
- Dependency review tools and package reputation services in CI.
Evidence an auditor would accept
- CI configuration and flagged examples.
Common mistakes
- Checking known vulnerabilities only, which misses brand-new malicious packages.
Not sure which requirements apply to your system? Take the assessment.
Requirement text from the OWASP AI Security Verification Standard, licensed under CC BY-SA 4.0. This guidance is shared under the same licence.