Twitter Will Now Prompt Users When They Reply With Offensive Language

Twitter is introducing prompts that lets users know when they use strong language and lets them review their tweets before posting them online. This new feature will be available for both Android and iOS users and for users who use the social media platform in English. The feature works when users tweet offensive remarks, insults, or strong language that could potentially be offensive or harm someone. According to Twitter, the social media platform tested out the feature in 2020.

Announcing its new feature through a blog post, Twitter explained the move comes to curb online hate among users. The social media giant will give users the opportunity to revise their tweets if it contains strong language that might incite hatred or be offensive to others on the platform. However, during their initial testing phase, users were prompted unnecessarily due to their algorithm. Twitter claims that it has now improved the algorithm to understand when a user is posting offensive tweets and when they are participating in friendly banter.

“Throughout the experiment process, we analysed results, collected feedback from the public, and worked to address our errors, including detection inconsistencies,” states the blog post. Through the testing process, Twitter said that it learned that when prompted, 34 percent of the users revised their initial reply or decided against posting the same. After being prompted once, people composed 11 percent fewer offensive replies in the future. Additionally, when prompted, people received fewer offensive and harmful replies.

After the testing phase, Twitter will identify the relationship between the author and the responder, including their frequency of interaction. The social media giant will also revise its technology to better detect the use of strong language. Additionally, Twitter has created an easier way for users to provide feedback regarding the usefulness of the prompts.