Physical Address

304 North Cardinal St.
Dorchester Center, MA 02124

Safety mode gives users a way to turn off harassment

 Hate speech and harassment will be blocked for seven days, as part of the Safety Mode functionality.

Once turned on, the feature will function automatically, sparing users the stress of having to handle unpleasant tweets.

It will be tested on a few customers first.

When turned on, the feature evaluates both the content of the tweet and the connection between the tweet author and the replier. Users are less likely to get autoblocked if they frequently interact with or follow accounts on Twitter.

Katy Minshall, who heads Twitter UK Public Policy, admitted that “there is always more to be done” to ensure user safety on the platform.

“We’re announcing Safety Mode, which provides an option to automatically moderate offensive or hurtful remarks on Twitter to make the public conversation more healthy.”

The main form of moderation used by Twitter is the same as that used by other social media sites, namely a combination of human and computer control.

Despite the fact that it has never given any numbers, New York business school NYU Stern reported in 2020 that the company had roughly 1,500 human moderators to handle the 199 million daily Twitter users.

According to a report published by the Finnish government, Twitter “does the worst of the tech giants in eliminating hate speech on its platform.”

In their research, the authors (Paukkeri) have found that artificial intelligence systems trained by humans are what is needed to address the challenge.

“There are a lot of ways to say something bad, and it’s like rocket science to create tools that can detect the terrible things,” she added.

“Simply highlighting some words or phrases,” she added, “is not enough.”

As well as working to prevent abusive content, Twitter is also trying to more aggressively target fake news. It teamed up with Reuters and the Associated Press in August to fight fake news, debunk lies, and stop them from spreading.

Prior to the introduction of the Birdwatch community moderation system, it allowed volunteers to label inaccurate tweets.

Leave a Reply