PII Detection Trend(5.4.7)

The absolute count of detections can be misleading if overall Jira activity is also changing. The detection rate, detections as a percentage of total events, is the more meaningful metric. A falling rate even as activity grows indicates that data hygiene is improving. A rising rate indicates that a growing proportion of work is generating personal data exposure.

What you can conclude

• A declining rate after a training initiative or process change shows when the rate fell; whether that change caused it is not in the data.
• A rising rate indicates that whatever is generating the detections is increasing: new integrations, new workflows, or team growth outpacing awareness.
• A stable rate indicates a steady-state pattern, with consistent and identifiable sources.

How this chart works

Weekly chart: each week's detections are red bars against the left axis, and the detection rate, detections as a percentage of all events, is a green line against the right axis. The right axis is scaled to the rates in view rather than fixed at 0 to 100%, so read its values before calling a move large: a change of a fraction of a percentage point can look steep. The week's total event count is shown in the tooltip rather than drawn. Use the date and project filters to focus on specific teams or periods.

What the detection count contains. Each week's detections union two arms: text matching an email address pattern, and text carrying a standalone run of 7 to 15 digits that does not parse as a date or a timestamp. Digits that are part of an Atlassian account id, such as the one behind a mention of a colleague, or of an identifier in the standard UUID format, such as the id of a file embedded in a comment, are not counted. Nor are the digits Jira writes into a comment's own formatting: the file name it records for an embedded image or file (a pasted screenshot is named like image-20260421-185001.png), the editor's internal element ids, and the stored value of a date inserted with the date picker. Digits in a link address still count. The second arm is much looser than the first, and on a portfolio with many long internal reference numbers it can dominate the count. Read a move here together with the PII Type Distribution visual, which shows which arm moved.

A falling trend is what a GDPR reviewer typically wants to see after a remediation effort.

The week in progress is shown at its value so far. Because the week is not yet complete, it covers only the events recorded so far and can move in either direction as the week closes.