You are using an out of date browser. It may not display this or other websites correctly. You should upgrade or use an alternative browser.
purview labels
About this tag
Purview labels are Microsoft's sensitivity labels used to classify and protect data in Microsoft 365. Recent discussions on WindowsForum highlight a significant privacy flaw (CW1226324) where Microsoft 365 Copilot Chat bypassed Purview sensitivity labels and Data Loss Prevention (DLP) controls, reading and summarizing emails marked as Confidential. This bug exposed a gap in enterprise data governance, as Copilot accessed labeled emails from Sent Items and Drafts despite configured protections. The issue was patched with a server-side update, but it underscores the need for careful governance of AI features within Microsoft 365. Topics include data classification, DLP, and AI compliance.
Microsoft’s flagship productivity AI for Microsoft 365 has a glaring privacy problem: for weeks a code error allowed Copilot Chat to read and summarize emails that organizations had explicitly labelled as confidential, bypassing Data Loss Prevention (DLP) controls and undermining a core tenant...
Microsoft has confirmed a logic error in Microsoft 365 Copilot Chat that briefly allowed the assistant to read and summarise email messages organizations had explicitly marked as Confidential, bypassing Purview sensitivity labels and configured Data Loss Prevention (DLP) controls — a lapse...
Microsoft’s Copilot Chat quietly summarized emails labeled “Confidential,” bypassing the data‑loss protections administrators relied on and forcing a hard assessment of how AI features must be governed inside Microsoft 365...