Twitter says some posts by politicians could get tagged with a warning

Presidents and other world leaders and political figures who use Twitter to threaten or abuse others could find their tweets slapped with warning labels.

The new policy, announced by the company on Thursday, comes amid complaints from activists and others that U.S. President Donald Trump has gotten a free pass from Twitter to attack his enemies in ways they say could lead to violence.

Starting Thursday, tweets that Twitter deems in the public interest but which violate the service's rules will be obscured by a warning explaining the violation. Users will have to tap through the warning to see the underlying message.

Twitter said the policy applies to all government officials, candidates and similar public figures with more than 100,000 followers.

The company has long exempted public figures from many of those rules, contending that publishing controversial tweets from politicians encourages discussion and helps hold leaders accountable.

There have been longstanding calls to remove Trump from the service for what activists and others have said is abusive and threatening behaviour.

Trump drew fire for a 2018 retweet of an Islamophobic British account, and some activists complained this week after the president posted that an attack by Iran "will be met with great and overwhelming force. In some areas, overwhelming will mean obliteration." 

In 2018, French prosecutors filed preliminary charges against far-right French politician Marine Le Pen for tweeting brutal images of Islamic State violence. Twitter rules prohibit material "that is excessively gory."

And in March, Brazilian President Jair Bolsonaro stirred outrage by sharing a video on Twitter that showed a man urinating on the head of another man during a Carnival party, although an explicit warning did eventually appear hours later.

Old tweets need not apply

Twitter's new policy doesn't apply to past tweets.

Twitter said it is still possible for a government official or other figure to tweet something so egregious that it warrants removal. A direct threat of violence against an individual, for instance, would qualify.

The company said warning-label decisions will be made by a group that includes members of its trust and safety, legal and public policy teams, as well as employees in the regions where particular tweets originate.

When a tweet gets such a notice placed on it, Twitter said it will not appear in the site's "safe search" mode, notifications tab, the explore section and other places Twitter promotes or highlights tweets.

The policy could anger some Republicans and conservatives, who have argued that the Silicon Valley social media giants are biased against right-of-centre viewpoints.

Trump has also complained about a "loss" of many Twitter followers in the past two years, a time period that coincides with the platform seeking to remove bots, or automated accounts, that violate its policy terms.