Twitter to beat trolls by hiding messages from other users

TWITTER is turning to a new tactic in the company’s long-running battle against internet trolls, by simply hiding their posts from view.

The social network, which has been accused of failing to delete abusive messages, will not display posts from users who exhibit “troll-like” characteristics.

This includes people who repeatedly post to people who do not follow them, those who have often been blocked by other users of the site, or those who set up multiple accounts at once.

Twitter says that while many posts do not break its rules, and so do not warrant users being banned, they may still be in poor taste or offensive.

While the posts will not be deleted, they will be far less visible, particularly in search results and when viewing replies to existing tweets.

By making the posts less visible, instead of having to make decisions on who to ban, Twitter hopes to maintain its free speech credentials while making its service more enjoyable for users.

However, it may lead to criticism that it is sweeping the problem under the carpet by refusing to take down offensive posts.

Del Harvey, the Twitter safety director, said less than 1 per cent of all accounts made up the majority of reports, and that many of these were “disrupting the conversation” rather than explicitly breaking rules.

“There were a number of accounts that didn’t violate our rules, people just disagreed with them,” she said, adding that the algorithm being based on people blocking and muting users meant Twitter’s users would be influencing what constitutes improper behaviour.

Ms Harvey said that in the markets where it had been tested, there had been an 8 per cent drop in reports from conversations and 4 per cent from searches.