Update 3/10/2021: 

A Twitch spokesperson told Inven Global: "We are exploring ways to improve the experience on Twitch for viewers and creators, including efforts to better match the appropriate ads to the right communities. User privacy is critical on Twitch, and, as we refine this process, we will not pursue plans that compromise that priority. Nothing has launched yet, no personal information was shared, and we will keep our community informed of any updates along the way."

Original Story:

Cyber Security student Daylam Tayari released screenshots of Twitch's public API on Tuesday that appear to show a new automatic "Brand Safety Score" for every streamer. Based on its name, the new score system will presumably be used by potential advertisers and sponsors to decide what channels they want to advertise on.

According to Tayari, the Brand Safety Score is based on a list of factors that includes:

• Age of the streamer
• Ban history
• Partnership status
• whether or not they use Automod
• The ESRB rating of the games they play
• Stream maturity settings
• Manual changes to the score by Twitch staff

While Twitch has not yet commented on the discovery of the new score system, the system is already facing backlash online. A number of people responded to the original post, voicing their concerns that the new score would be used to punish mature content. Many also took issue with Twitch's ability to manually alter scores, making the system more arbitrary.

"This gives Twitch more power over creators." Said Twitter user XiloTLDR. "I don't like the platform being in charge of the score. What if it's used to gatekeep? What if this creates a power system that bad actors can use to exploit streamers or brands? What are the possible unintended consequences?"

The fears expressed in the comment section are not unwarranted. Over the past few years, YouTube has increased its regulation of mature content, in many cases demonetizing or restricting content from creators who touch on mature or controversial topics. Making matters worse, YouTube has also been accused of playing favorites when it comes to enforcement of these policies, with the platform punishing some creators while ignoring questionable content from their heavy performers. 

It appears that Twitch maybe implementing a brand safety system similar to YouTube's, and like YouTube, Twitch has been accused of playing favorites in the past when it comes to how they hand out bans. The new Brand Safety Score would expand the risk of favoritism further since Twitch staff can manually adjust scores. Even if a system like this would be good in theory, many folks don't trust Twitch to implement it fairly.