Introducing Topic Detection
Automatically detect and act on content about sensitive topics like religion, politics, and SHAFT categories.
Not all content that needs attention is harmful. Sometimes you need to find and manage content about specific subjects - whether for compliance, policy enforcement, or keeping conversations on track.
Today we're launching topic detection, making it easy to identify and act on content about specific subjects.
What you can detect
Topic detection currently covers two categories:
Sensitive topics
Flag content discussing religion or politics. Useful for platforms that want to keep spaces neutral or route potentially heated discussions to specialized reviewers.
Illicit content & SHAFT
Detect content related to Sex, Hate, Alcohol, Firearms, Tobacco, Drugs, and Gambling. If you're sending SMS through telecom providers, you likely already know these categories - many carriers require SHAFT compliance.
We also cover general illicit content for broader policy enforcement.
How to use it
Enable topic detection from your channel settings. Select which topics you want to detect and they'll automatically be flagged on incoming content.
Once you've selected your topics you can try out the new configuration in the playground.
Route topics to dedicated queues
Combine topic detection with content queues to create focused review workflows. Set up a queue filtered to political content, and your trust & safety team sees only what they need to see.
Need a specific topic covered?
We're expanding topic coverage based on what platforms actually need. If you have a use case that requires detection of specific subjects - whether for compliance or community management - reach out and let's talk.