AI data and models
Choose which models VeriProof uses for its AI tasks. Only organization administrators can change these settings. Review the data each task sends before enabling it.
Enable AI features
If your organization has no saved settings here, available AI tasks use the deployment’s default configuration. The page identifies this state explicitly. Unsaved controls do not describe the active configuration.
After you save, both controls must be enabled for the listed tasks to use an AI model:
- Enable AI features permits VeriProof to run configured AI tasks.
- Allow approved data use permits VeriProof to send the data required by an enabled AI task to the configured service.
Activity analysis sends redacted session data. Disabling either control and saving prevents these tasks from sending data to an AI model. Template-based results remain available.
Review your organization’s data-use requirements before you enable either control. VeriProof does not send unredacted session data from this workflow.
Assign models
Review each task
Review narrative summaries, anomaly analysis, questions and answers, translation, and portfolio analysis separately. A model approved for one task may not suit another.
Select an approved model
Choose the model your organization approved for that task. The available list is controlled by the current VeriProof configuration.
Select the reasoning level
Choose a supported reasoning level for the selected model. After changing models, check this value again: VeriProof adjusts a level the new model does not support.
Save and verify
Select Save changes. Confirm that the page reports AI data and model settings saved before you leave.
If loading fails, do not treat missing values as the saved configuration. If saving fails, the change is not confirmed. Ask the administrator responsible for your deployment to review the AI service connection.