Think twice the next time you post something potentially rude or harmful in Twitch chat. The streaming site is starting to use AI to detect “offensive or disrespectful” language.

Twitch's support account on X (formerly Twitter) announced today that it will be experimenting with a new feature "over the coming weeks" that attempts to detect potentially negative posts in order to filter out the bad vibes — and potentially prevent you from getting banned.

"We're rolling out an experimental prompt powered by machine learning that encourages users to pause before sending potentially harmful chat messages," Twitch said. "Our goal with this experiment is to reduce harassment and facilitate more positive interactions on Twitch."

The vague example Twitch provides shows a message prompt that appears when you try to send the message in question, asking “are you sure you want to send this?” It warns that “your message may contain words that may be considered offensive or disrespectful” and gives the user the option to either opt out of the message or “send anyway.”