Twitter Developing New Feature to Filter Hate Speech (Report)

Jamie McCarthy/Getty Images

The site has been citicized for failing to curb harassment and abusive language.

Twitter is working on a new safety tool that prohibits specific keywords from appearing in news feeds, allowing users to filter out speech that they find hurtful or otherwise offensive, according to Bloomberg, which cites multiple unnamed sources. 

The micro-blogging site, which has referred to itself as the "free speech wing of the free speech party," is trying to figure out how exactly to implement the new filtering feature, which has been in development for about a year, the article reports.

The San Francisco-based tech company has long sought to strike a balance between unrestricted speech and personal safety. In 2015, then-CEO Dick Costolo wrote in an internal memo to employees:

"We suck at dealing with abuse and trolls on the platform, and we've sucked at it for years. It's no secret, and the rest of the world talks about it every day. We lose core user after core user by not addressing simple trolling issues that they face every day."

One core user who quit the site after a barrage of trolling was Saturday Night Live comedian Leslie Jones. She has since returned to the platform, but her personal website was hacked this week and defaced with derogatory language reminiscent of the original Twitter attack.

Twitter isn't alone in its desire to curb hate speech. Last year, Alphabet chairman Eric Schmidt called for hate speech "spell checkers" that would block harassment using similar, keyword-based approaches.

Twitter has reportedly already used these kinds of tools. In a BuzzFeed exposé from earlier this month, Charlie Warzel reported that Dick Costolo "secretly ordered the media partnerships team inside Twitter to use an algorithm to filter all tweets directed at [President Obama] for abusive language" during the question-and-answer portion of a May 2015 speech.