Microsoft has plans for early November 2026 the arrival in Teams AI-based phishing detection that analyzes conversation context to identify suspicious behavior. The release starts with a targeted release phase which should close in mid-November, while the general availability in the world it is expected, in all probability, between the beginning and mid-December. The feature appeared in mid-October in an entry on the Microsoft 365 roadmap and was confirmed in an update to the Message Center last week.
For most users, detection focuses on first contact external chatsone of the highest risk messaging scenarios. Microsoft cites phishing, social engineering, impersonation attempts and deceptive intent as targets. Context analysis serves to identify suspicious behavior even in messages without traditional indicators, such as Malicious links or files.
IT administrators will not have to do anything to activate protection. Microsoft states that the feature is for security purposes only: “This feature is used for security purposes only, and not for productivity, summaries, advertising, user tracking, or other non-security purposes.” The content of customer messages, the company adds, is not retained by the detection system for model training or other secondary uses.
The announcement adds to a series of actions by Microsoft against the use of Teams for social engineering attacks. In December the company announced that administrators would be able to block external users from the portal from January Defenderto hinder criminal groups, including ransomware, who exploit the platform to target employees of target companies. Since August, a meeting protection policy has also been distributed that allows everyone to be automatically blocked from entering external bots identified.
Last month Microsoft announced that Teams will blur i QR codes sent by external senders, to limit fraud and phishing. Users will also be able to report suspicious invitations from guests directly in Teams, so security teams can block attacks using this route. In early October, Microsoft announced that it would introduce support for third-party tracking solutions deepfake and protection from impersonation in Teams meetings.
