How to meet it
OWASP AISVS v1.01
Monitoring, Logging & Anomaly Detection · Model, Data, and Performance Drift Detection
Detect hallucinations
Verify that hallucination detection monitors identify and flag model outputs that contain factually incorrect, inconsistent, or fabricated information.
OWASP AISVS v1.01, requirement 12.3.2
In plain English
Monitor answers for factual errors, contradictions and made-up content, and flag them.
What good looks like
- A detection method runs on production answers or a sample of them.
- Flagged answers are logged and reviewed.
- The method is checked against known examples.
How teams usually do it
- Groundedness checks against retrieved sources, or a second model checking claims.
Evidence an auditor would accept
- Detector configuration and flagged samples.
Common mistakes
- Measuring hallucination only in pre-release testing.
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.