AI-Assisted Secure Coding (Appendix C) · AI-Assisted Secure-Coding Workflow
Name the AI threats you defend against
Verify that the workflow names the adversarial-AI threat scenarios it is built to mitigate. The list should cover prompt injection delivered through PR content, AI-generated supply-chain payloads, autonomous agents approving their own work, fork-PR secret exfiltration, and compromise of the model supply chain.
OWASP AISVS v1.01, requirement AC.1.3
In plain English
List the attacks against AI coding tools your workflow is meant to stop, such as prompt injection through pull requests, malicious generated dependencies, agents approving their own work, secret theft through forks and a compromised model supply chain.
What good looks like
- Each listed threat maps to at least one control.
- The list is reviewed when tools change.
- New incidents in the industry are checked against it.
How teams usually do it
- A threat model document for your AI-assisted development workflow.
Evidence an auditor would accept
- The threat list with mapped controls.
Common mistakes
- Listing threats with no link to the controls that address them.
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.