Toxicity in online games is no news. Lately, however, several companies have been trying to get it under control. In March 2018, we covered the foundation of the 'Fair Play Alliance', an open forum in the games industry to collaborate on best practices that would foster community growth and fair play.

　　Microsoft, which is part of the Fair Play Alliance, has now moved forward with the brand new 'message safety' text filters announced yesterday. The feature, which lets you customize four degrees of filtering, is rolling out to Xbox insiders today with the goal to release it to all users later this Fall.

　　It should be noted that once enabled the text filters will also be applied to the Xbox apps on Windows PC and mobile.

Text

　　This applies to both messages and message requests (that is, messages from people who aren’t already on your friends list). You've always been able to report offensive messages, now you have the opportunity to filter them out entirely.

　　FriendlyAt this level, our filter detects and blocks as much potentially offensive content as possible, without interfering with discussions of themes common to games and speech most people would consider normal. We’ll do our best to filter common vulgar words, bullying, anything related to sex, and any harmful content based on protected characteristics such as race, gender, and nationality.

　　We can’t promise to catch everything that could be harmful, but the Friendly setting is our best option for people wanting to block as much potentially offensive content as possible.

　　MediumThe Medium text filter aims for a balance between Friendly and Mature. This setting recognizes that people often use harsh words even when they’re not speaking with harmful intent. Accordingly, Medium will allow certain vulgar words so long as they’re not being used to bully someone, target protected characteristics (such as race, gender, and nationality), or reference sexual acts.

　　We recognize that everyone has their own opinions on what language is “bad” or “good.” While we develop more personalized safety features, we offer the Medium setting as a good starting point for people who want to be reasonably safe without overly filtering their experience.

　　MatureOur Mature text filter is lenient around vulgarity but will block content we identify as harsh bullying, explicitly sexual, or severely targeting protected characteristics such race, gender, and nationality.

　　The Mature setting is intended for people who mostly trust who they’re talking to and don’t mind harsh language from those players.

　　UnfilteredWith Unfiltered, our text filter is effectively off. We recommend this only for private conversations with friends you trust wholeheartedly.