Twitch's off-service conduct policy finally covers doxxing and swatting

During the TwitchCon Las Vegas Opening Ceremony on Friday, Twitch CEO Dan Clancy made an important announcement regarding the platform's off-service conduct policy. Starting now, Twitch will take action against users who doxx and swat others outside of its own platform.

Doxxing (publishing someone's personal information such as their full name and address) and swatting (making a prank call to dispatch emergency services to someone's home) have long been problems in the livestreaming community. Now, if Twitch catches a streamer or other user doxxing or swatting someone on another platform or even offline, it may suspend or ban their account.

Twitch started enforcing its off-service conduct policy in April 2021. At the time, it said it would respond to serious incidents that "pose a substantial safety risk to the Twitch community" even if those happen away from its own platform. Before it makes any enforcement actions related to off-platform activity, Twitch requires evidence of such incidents taking place.

The policy covers issues such as sexual misconduct, terrorist activity and recruitment, explicit or credible threats of mass violence and being a leader or member of a known hate group. The platform also takes action related to threats against Twitch and its staff, as well as those who share harmful misinformation. Adding doxxing and swatting to the list is a long-overdue move given the serious and occasionally even deadly nature of such actions.

Twitch had some other safety-related announcements to make at TwitchCon. Streamers and mods will soon be able to anonymously send warnings to chatters who step out of line. The chatter in question will need to acknowledge the warning before they can type in the chat box again. Twitch notes that many streamers have an informal three-strike system in place to keep their communities safe, and this channel-level chat warning feature aligns with that.

Earlier this year, Twitch started testing an AutoMod feature called Smart Detection. This tweaks AutoMod (a tool that automatically keeps inappropriate content out of stream chats) based on the moderation actions that streamers and their mods make. Smart Detection is now live on all English-language channels. Twitch plans to roll it out more broadly in the future.