When PII detection is on, DataMind OS checks chat messages for personal data before they are sent to the AI. In a chat you can scan a draft while you write, and if a message is flagged you are shown what was found so you can fix it.
email · 92%. On a touch screen the tooltips do not appear.A check can also open the full result in a PII scan dialog, for example when you check a training card. The scan starts as soon as the dialog opens, and Scanning… shows while it runs. There is no separate scan button inside; the dialog says the text is checked against the platform's PII detection policy.
email - 2.. * instead of a percentage.What happens next depends on the policy:
The check runs the saved PII detection policy for what you are checking: the same rules and engines used when that text is sent or saved. Changes that have not been saved yet do not apply. The policy is set on Admin > PII Detection; see PII detection.
Every detection and every skipped warning is recorded in the PII log; see AI, socket and PII logs.
It is hidden while Global PII detection is off, because no checks run anywhere then. An administrator can turn it on in Admin > PII Detection. See PII detection.
The check needs something to read. It is unavailable while the draft is empty or holds only spaces, while a check is already running, or while the message box itself is disabled. Wait for the running check to finish, or type some text first.
No. The check reads the draft and colours what it finds. Nothing is posted to the chat until you send the message yourself.
The saved PII detection policy for chats: the same rules and engines used when a message is sent. Edits an administrator has made but not saved yet do not apply. See PII detection.
The colours follow your text. A highlighted part whose own text you change loses its colour, because it no longer matches what was found. The parts you did not touch keep theirs and shift with your edit. Run Check PII again for a fresh result.
. * instead of a percentage?The match came from a regex rule. A regex either matches or it does not, so there is no confidence score to show.
Only in Warn User mode and only if you hold PII Detection Bypass: choose Send it anyway on the notice or in the dialog. It applies to the exact text that was flagged. In Block User Traffic mode the message cannot be sent until the personal data is removed or masked.
Three reasons: the policy is Block User Traffic, so the block is final; you do not hold PII Detection Bypass; or you edited the message after it was flagged. An edit cancels the bypass for that text, and the edited message is checked again when you send. Without the permission the option is not offered and a bypass send is refused. See the permission reference for what each PII permission covers.
The tooltips that name a label and its confidence score are for pointer devices; on a touch screen they do not appear. The colours, the scan result and the notices still show.
Choose Restore on the notice, or Edit message in the dialog, and the text returns to the message box. The notice confirms "Your message is back in the input." It stays on screen for 12 seconds, so act while it is visible.
Yes. Every detection and every skipped warning is recorded in Observability > Logs > PII, and logging cannot be turned off. Reading the entries needs PII Log Viewer. See AI, socket and PII logs.
The check could not run. The notice shows the reason the server gave; if there is none, it reads PII scan failed. Run Check PII again, and if it keeps failing, tell your administrator and quote the message.