(Reuters) – Twitter Inc said on Thursday it would label or remove misinformation aiming to undermine confidence in the U.S. election, including posts claiming victory before results have been certified or inciting unlawful conduct to prevent a peaceful transfer of power.
Twitter said in a blog post it was updating its rules to recognize the changes in how people will vote in the Nov. 3 election and try to protect against voter suppression and misleading content on its platform.
The widespread use of mail ballots in the U.S. election due to the coronavirus pandemic will likely cause significant delays in tallying results, which some experts fear could allow misinformation to gain traction.
U.S. President Donald Trump has repeatedly claimed, without evidence, that voting by mail is susceptible to large-scale fraud.
Twitter also said it would label or remove misinformation creating confusion about the laws, regulations and officials involved in civic processes, as well as disputed claims that could undermine faith in the process, such as unverified information about vote tallying or election rigging.
A Twitter spokesman said whether content had specific falsehoods or could cause greater harm would determine if it would be removed, or labeled and have its reach reduced.
Social media companies have long been under pressure to combat misinformation after U.S. intelligence agencies determined Russia used their platforms to meddle in the 2016 presidential election, allegations Moscow has denied.
The companies have also been under scrutiny over their responses to inflammatory content posted by Trump. Since May, Twitter has attached warnings and fact-checking labels to Trump’s tweets about mail-in ballots.
Twitter said its rules would be “applied equally and judiciously for everyone.” The new policy, which is global, will take effect on Sept. 17.
Alphabet Inc’s Google also said on Thursday it would remove the function that attempts to predict and complete search terms when people look up the status of voting locations, voting requirements or methods – for example, “you can vote by phone” or “you can’t vote by phone” – though users will still be able to search for this information.
It will also remove those autocomplete predictions in searches about the integrity of the election, and claims for or against any candidate or political party.
Google staff told reporters on a call that incorrect information about election results, including reports claiming an early victory, would not show up on Google Search and that it would enforce its ads policy against demonstrably false claims that could undermine trust or participation in an election, including in the post-election period.
Facebook Inc last week said it was creating a label for posts by candidates or campaigns that made premature claims of victory. It also said it would stop accepting new political ads in the week before Election Day.
(Reporting by Elizabeth Culliford; Editing by Steve Orlofsky and Bill Berkrot)
