AI Risk Remediation

​

Remediation Advice

​

Validating Remediation

After attempting remediation, it is important to re-test the target API to ensure that the risk has reduced as expected.

If you have set up a CI/CD pipeline integration then the remediation action should automatically trigger a re-test via the Mindgard CLI.

If you are testing via the Web UI you can schedule a new test to validate your remediation via the Test Configurations page.

After running the test you can validate that the risk has reduced and how effective your remediation was.

This illustration shows an increase in risk as a system prompt is changed to be less effective, followed by a reduction in risk as it is improved.