Twitter Spaces filled with hate speech, extremism

Twitter's live audio chat, Spaces, is struggling to contain harmful content, including hate speech and extremism instigated by Taliban supporters, white-nationalists, and anti-vaccine activists. The Washington Post reported evidence of the content present on the platform which involved derogatory comments on Shia Muslims, black Americans, and transgenders. Hundreds of people were listening in on these live chats spreading hate.

Twitter users can report live audio chats, flagging content that is inappropriate. The company reviews the flagged content after saving the video for 30 days after its reporting. The social media platform has AI software that detects offensive words in the Space's titles. It does not have human moderators appointed that will review content on the platform in real-time.

Company employees report that they have to be careful around testing products for safety before they are released. Twitter's software also once accidentally promoted harmful live audio chats.

A spokeswoman for Twitter told Cnet that the company discovered bugs in its software which prolonged the time it took for Twitter to remove reported content. She said, "This left harmful conversations prominently visible and was a mistake."

Twitter's Spaces was tested out in November 2020 which is a social audio app, like Clubhouse.