Twitter Testing The Safe Mode Feature To Calm Abuse
Twitter is launching a feature it hopes will help tackle abuse and phishing, which has become a major concern for the platform.
Safe Mode will highlight accounts that use hate speech or bombard people with unwanted comments and block them for seven days.
The feature will work automatically once enabled, eliminating the task of users dealing with unwanted tweets.
First, it will be tested on a small group of users.
The function can be activated in the settings, and the system will evaluate the content of the tweet and the relationship between the tweet author and the reply. Accounts that are followed by the user or with which the user frequently interacts will not be automatically blocked.
Katy Minshall, Head of Public Policy for Twitter UK, said: “While we have taken steps to give people more control over their Twitter security experience, there is always more to do."
“We are introducing the safety mode; a feature that allows you to automatically reduce disruptive interactions on Twitter, which in turn improves the health of the public conversation."
Like other social platforms, Twitter relies on a combination of automatic and human moderation.
Although it never officially said how many human moderates it uses, a 2020 report from NYU Stern's New York Business School suggested it has about 1,500 to deal with the 199 million users on Twitter dailies worldwide.
A recent hate speech study conducted by Facts Against Hate on behalf of the Finnish government found Twitter to be the "worst tech giant" for hate speech.
The answer, according to study author Dr. Mari-Sanna Paukkeri, is to use artificial intelligence systems trained by humans.
"There are so many different ways to say bad things and it's rocket science to build tools that can identify them," he said.
It wasn't enough to distract specific words or phrases, a technique on which many social networks are based, he added.
In addition to dealing with platform abuse, Twitter has become more determined to tackle deformation. In August, it partnered with Reuters and the Associated Press to dismantle the misleading information and prevent its spread.
He previously introduced Birdwatch, a community moderation system that allowed volunteers to highlight abusive tweets.