Employees will receive another warning layer for scams that use conversation rather than malicious links or attachments. Microsoft is preparing AI-based phishing detection for Teams, focused mainly on first chats with external contacts. Targeted Release is expected in early November, with worldwide general availability beginning in early December. Detection may initially run silently while Microsoft validates quality.
Why It Matters
The feature is designed to identify impersonation, social engineering and deceptive intent from conversational context. For leaders: collaboration platforms now require the same social-engineering attention as email. For defenders: teach users to verify unexpected support contacts through a trusted channel and report suspicious chats. No new administrator configuration is required, and Microsoft says message content is not retained by this detection system for model training or secondary uses. The larger lesson is that the absence of a warning never proves a conversation is safe.
What Defenders Should Do
Update phishing training to cover external Teams chats and require independent verification of unexpected support requests.