Twitter has said it will not allow any of its users – including candidates – to claim victory prematurely in the forthcoming US election.
It said any tweets claiming victory before a clear projection would have a warning label added.
The move follows a similar decision from Facebook last month.
Twitter will also temporarily change the way retweets work – prompting users to “quote tweet” and add their own thoughts before spreading a message.
On top of its ban on “premature” claims of victory, Twitter said it would also remove any tweets that incited interference with election results, “such as through violent action”.
“Under this policy, we will label Tweets that falsely claim a win for any candidate and will remove Tweets that encourage violence or call for people to interfere with election results or the smooth operation of polling places,” the company said in a blog post.
Twitter gave election results as an example of one area covered by its new civic policies, which were updated in September. But Friday’s post sets a high bar for what will be allowed.
The benchmark for when Twitter users can claim a result will be either an announcement from state election officials or the projection of two “authoritative”, independent national news outlets.
Facebook boss Mark Zuckerberg has publicly addressed the problem, writing: “With our nation so divided and election results potentially taking days or even weeks to be finalised, there could be an increased risk of civil unrest across the country.”
As well as warning users over results, Twitter announced a handful of extra election-related measures on Friday.
Tweets that have had warning labels added to them will now see a pop-up prompt if they attempt to re-tweet them, with a prominent link to “credible” information about the topic.
Tweets from major accounts – such as US politicians or US-based accounts with more than 100,000 followers – will also only have the option to “quote tweet” on labelled tweets. And the social network said it wouldn’t recommend such tweets to new people through its algorithms.
It also said it would halt recommendations from strangers in timelines, and restrict recommended trends for US users to topics that include “additional context” – a description or article that explains the trend.