Microsoft Introduces New Protection in Teams: AI to Identify Potentially Harmful Messages

The new feature will use AI to analyze the context of conversations, including the way an unknown person addresses the user and what they ask for. This will help identify attempts at phishing, impersonation, and other forms of manipulation. The system will not just look for suspicious links or attachments, but also try to detect malicious intentions hidden in the messages themselves.

Microsoft has confirmed that the new feature will start rolling out in early November 2026, with the first phase expected to be completed by mid-month. The wider availability to users worldwide is expected between early and mid-December. The protection will be available in the computer, mobile, and web versions of Teams.