Harassment Filter

Filter content likely to contain abuse and harassment

Targeting Set based on the level of filtering
Content action Choose to review or remove comments
Allow list Ignore certain keywords. For partial matching, use * at the beginning, end, or both of each word. e.g. cat* or *cat*.

0/15 terms

Filter mod mail messages Note: Allow list doesn't apply to the mod mail filter

Frequently asked questions

How does the filter work?

The filter scores comments for likely abuse or harassment and takes action based on your settings above. Matches can either be routed to your mod queue for review or removed automatically.

What counts as harassment?

The model is trained to detect content that targets a person or group with insults, threats, or demeaning language. It doesn't catch every case, which is why you can combine it with the allow list and your own mod queue workflow.

Does the allow list apply everywhere?

No. The allow list is applied to comments on posts in your community. It does not apply to mod mail — the mod mail filter runs independently.

Learn more about the harassment filter

Filter comments