We are improving AI Reviewer flexibility by allowing partners to define how checks should behave when users provide missing or placeholder values instead of valid data.
What’s new?
AI Reviewer configuration now includes a new setting: “If the required information is empty or marked as ‘None’, consider task outcome as”. Available outcomes:
- Accepted
- Unaccepted
- Skipped
Key updates:
- Smarter Validation Logic: AI Reviewer can now recognize empty, placeholder, or non-validatable values such as “NONE”, “N/A”, “NULL”, “-”, and similar inputs.
- Reduced False Positives: Partners can avoid unnecessary “Flag for investigation” outcomes when certain data, such as VAT or EIN numbers, is intentionally unavailable.
- More Flexible Rule Configuration: Checks can now be skipped instead of automatically failing when required information is missing or not applicable.
- Backward Compatibility Preserved: Existing AI Reviewer steps without this configuration will continue working as before.
This update helps reduce incorrect investigation flags and provides partners with more accurate and configurable AI-powered review flows.
