Twitter is developing a new tool to prevent abuse by allowing users to filter out offensive tweets using keywords.
The social media company has been debating how to use the tool for roughly a year, Bloomberg News reports. According to Bloomberg’s sources, the tool is not yet public.
Twitter declined to comment on the report to The Hill.
Twitter has long been criticized for how it handles abuse on the platform. High-profile users, like actress Leslie Jones and British Member of Parliament Jess Phillips, have left or considered leaving the platform after enduring a deluge of slurs, threats and other discriminatory attacks.
Twitter has acknowledged these problems and made attempts to mitigate instances of abuse. Last Thursday, it announced a new “quality filter” feature that helps reduce the presence of automated and spam accounts.
Others have speculated about a feature to filter offensive tweets.
In 2014, The Verge reported on a story in which British Member of Parliament Luciana Berger had become the target of heavy abuse on Twitter. The abuse didn’t stop and the accounts of her harassers were not being deleted. According to The Verge, people were still tweeting offensive messages at Berger, but she wasn’t receiving them. Tweets that contained certain slurs directed at Berger’s account would be met with an error message upon sending.
The tech news website speculated that Twitter was utilizing a tool that hadn't been made public, similar to the one described in the Bloomberg report, to stop users from harassing the British lawmaker.
The tool, while potentially very useful in preventing abuse, could also be used to filter other things. Users could potentially filter out conversations on matters or events that didn't interest of offended them.
Facebook has a similar tool available for business users of Instagram that allows them to block comments containing words they don't want to see.
In the case of Luciana Berger though, these measures weren't perfect. The Verge notes that when abusers caught on to certain slurs being filtered, they placed dashes in between the letters of offensive words or tweeted pictures at her to bypass the safeguards.