Event banner
Data security: Zero Trust in a world of AI and emerging technology
Event details
That concludes our Unpacking Endpoint Management for today. We’ll be back next month at MVPs on manageability: reflections from the MVP Summit!
In addition to the questions posted on this page, we also answer questions posted in reply to the event on LinkedIn and X (Twitter). Here are the questions we answered today:
Question -- How can AI help with my unstructured data? We have stuff all over the place – even on individual computers - answered at 3:10
From X -- As a security pro, AI is my worst nightmare. Make me feel better about it? What are the 3 things I should get in line before my LT says "Go make AI happen!"? - answered at 6:50
Question -- We use WIP (Windows Information Protection) for Windows BYOD information protection. What should we be using now that it has been deprecated? - answered at 13:20
Question -- Are there differences between MIP unified labeling client and labels built into Office apps? - answered at 14:50
From LinkedIn -- Can I stop data exfiltration before it actually happens? - answered at 16:15
Question -- Are encrypted files encrypted both “at-rest” and “in-transit?” - answered at 22:40
From X -- You mentioned AI Hub. Where can I find more information about that? How do I get access? - answered at 29:10
From LinkedIn -- I think I heard about a new Purview compliance feature that could help me analyze messages across Copilot, Teams, email, etc. more easily and assess risk from the inside. Can you talk more about that? (Wish I could remember the name of it...) - answered at 33:20
Question -- What about PDF files. Especially password-protected files?. Can we protect sensitive information in those file types? - answered at 40:10
Question -- What's the best way to establish uniform protection policies across multiple data sources? - answered at 48:55