Threat, profanity, and harassment trainable classifiers available in new languages

Microsoft 365 compliance solutions provide machine learning classification through trainable classifiers.

Built-in models for threat, profanity, and harassment can now support detection for seven languages, in addition to English, in a preview:

  1. Chinese (Simplified)
  2. French
  3. German
  4. Italian
  5. Japanese
  6. Portuguese
  7. Spanish

This message is associated with Microsoft 365 Roadmap ID 70623

When this will happen

  • Public preview: late March through mid-April

How this will affect your organization

If you have an existing policy which utilizes threat, profanity, and/or harassment trainable classifiers, any matches found in English as well as in the seven newly supported languages will be detected. 

What you need to do to prepare

There is no action required for existing policies using threat, profanity, and harassment to start picking up matches in new languages.

If you do not have a policy created using these classifiers and would like to detect threat, profanity, and harassment in a newly supported language, you can create one in available solutions.

Message ID: MC246740


No comments yet

Leave a Reply


I've been working with Microsoft Technologies over the last ten years, mainly focused on creating collaboration and productivity solutions that drive the adoption of Microsoft Modern Workplace.

%d bloggers like this: