Twitch has introduced a Suspicious Account Detection feature to assist and protect streamers as well as their conversations from abuse. The technology is being developed in response to hate crimes and abuse on the platform, particularly among streamers from disadvantaged populations. It tries to capture users avoiding channel-level bans by identifying potentially suspect conduct in chats, which are driven by machine learning. When users ban somebody from their channel, they should be permanently barred from your community. Sadly, bad actors frequently establish new accounts, return to the chat, and resume their distasteful activities. Suspicious User Detection, backed by machine learning, is built to assist users in identifying such gamers based on a variety of account indications. There are 2 tiers of suspicious behavior flags. Anything message in chat for “probable” ban avoiders will be accessible only by broadcasters and their moderators.
A notification will be displayed, but, broadcasters and moderators will be alerted of “potential” ban evaders. Messages can then be deleted and users blacklisted. This utility will be enabled by default, across all streams. Streamers will be able to turn it off or change the settings to be more careful, such as prohibiting any communications identified by the program. One thing to keep in mind though, particularly as the launch date approaches, is that hardly any machine learning system will ever be completely correct, which is why prohibitions on suspicious activity aren’t automated. When it comes to their community, users are the experts, and users should make the final decision on who may/may not join. The technology will learn from your behaviors, and the accuracy of its forecasts should increase with time.
This tool exemplifies our general approach to safeguarding technology: create strong tools that function in tandem to provide you with greater control over your community. Our job is never done, and we’re always looking for new ways to combat hatred, abuse, and ban avoidance on Twitch. The new technology has received strong feedback on Twitter, as consumers claim it represents a huge step forward for the streaming platform in keeping its users safe.