You are not seeing successful detection in chat content, for the Microsoft Copilot Gatelet you have configured.
How can CASB help protect sensitive information in Copilot where sensitive information could be leaked to an employee based on a prompt from the employee.
CASB with a Custom Gatelet configured for Microsoft Copilot
For Gatelets: Feature Request "ISFR-3036: Gatelet for Microsoft Copilot" has been created requesting full gatelet functionality.
For Securlets: (API): Securlet is not yet possible without further Microsoft development of an API.
Microsoft has suggested one of the following approaches to restrict sensitive information in response to users' prompts. In other words, preventing the accidental leakage of sensitive information by Copilot to an employee as a response to a prompt from the employee.
Approach 1 "Restricted SharePoint Search"
Restricted SharePoint Search allows admins to disable organization-wide search, and to restrict Copilot to a curated set of SharePoint sites. Admins can set up an allowed list of curated SharePoint sites.
Additional Documentation:
Approach 2 "BlockContentAnalysisService"
This approach uses Microsoft sensitivity labels with the new BlockContentAnalysisService setting to restrict the content. The label can be applied to content automatically through a DLP incident were the response action sets the MPIP label. The content would need to be labeled before the BlockContentAnalysisService applies.
Additional documentation:
Block Copilot Access to Individual Office Documents
Symantec DLP / MIP Integration
Remediation with Microsoft Purview MPIP using the o365 Securlet and DLP Enforce
Protecting data through copilot.microsoft.com
Bing has copilot. Uploads were blocked, chats were not blocked.