AI-Assisted Secure Coding (Appendix C) · Compromise Containment & Automated Remediation
A playbook for AI pipeline compromise
Verify that an incident-response playbook exists for AI-in-pipeline compromise. At minimum it covers: revoking AI-agent credentials, rotating every secret that touched the compromised workflow run, quarantining the compromised artifacts, notifying downstream consumers, notifying regulators where applicable, and preserving prompts, responses, and audit logs for forensics.
OWASP AISVS v1.01, requirement AC.14.1
In plain English
Have an incident response playbook for a compromise involving AI in your pipeline. It covers revoking agent credentials, rotating every secret the run touched, quarantining artifacts, telling downstream users and regulators where needed, and keeping prompts, responses and logs for investigation.
What good looks like
- The playbook exists and covers each listed step.
- Owners are named.
- It is linked from your main incident process.
How teams usually do it
- Extending your existing incident playbooks with an AI pipeline scenario.
Evidence an auditor would accept
- The playbook and its last review.
Common mistakes
- Forgetting to preserve prompts and responses before rotating and cleaning up.
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.