Twitter labels Trump's false claims with warning for first time

Julia Carrie Wong in San Francisco and Sam Levine in New York
Photograph: Jonathan Ernst/Reuters

Twitter has for the first time taken action against a series of tweets by Donald Trump, labeling them with a warning sign and providing a link to further information.

Since ascending to the US presidency, Trump has used his Twitter account to threaten a world leader with war, amplify racist misinformation by British hate figures and, as recently as Tuesday morning, spread a lie about the 2001 death of a congressional aide in order to smear a cable news pundit. Throughout it all, Twitter has remained steadfast in its refusal to censor the head of state, even going so far as to write a new policy to allow itself to leave up tweets by “world leaders” that violate its rules.

The company’s decision on Tuesday afternoon to affix labels to a series of Trump tweets about California’s election planning is the result of a new policy debuted on 11 May. They were applied – hours after the tweets initially went out – because Trump’s tweets violated Twitter’s “civic integrity policy”, a company spokeswoman confirmed, which bars users from “manipulating or interfering in elections or other civic processes”, such as by posting misleading information that could dissuade people from participating in an election.

Trump responded on Tuesday evening with a pair of tweets that repeated his false claims about voting and accused Twitter of “interfering in the 2020 Presidential Election”. “Twitter is completely stifling FREE SPEECH, and I, as President, will not allow it to happen!” he wrote. Federal law protects the rights of internet platforms to moderate the third-party speech they publish.

Trump’s tweets include numerous false statements about California’s plan to expand access to voting by mail in November due to the coronavirus outbreak. The tweets now feature a light blue exclamation point icon, with the message “Get the facts about mail-in ballots”. The alert label may not be visible when the tweets are embedded in another web page, such as below.

Clicking on the alert will link users to a “Twitter-curated page” that variously describes Trump’s claims as “unsubstantiated” and false. The Twitter page also aggregates tweets from a number of journalists and publications explaining why Trump’s statements are false.

Trump’s claims about California on Tuesday were blatantly wrong. The state is not sending a ballot to anyone who lives in the state but rather those registered there. Sam Mahood, a spokesman for the secretary of state, Alex Padilla, said in an email only active voters in the state would be mailed ballots.

As states prepare for an unprecedented surge in mail-in voting because of the Covid-19 pandemic, Trump has repeatedly made baseless claims that this will lead to fraud. Voter fraud is extremely rare and one analysis found just 143 convictions involving mail-in ballot fraud since 2000, representing 0.00006% of the ballots cast during that time period.

Trump, who voted by mail in Florida in March, has made it clear that he opposes any effort to make it easier to vote by mail for all eligible voters, including sending an absentee ballot application to all voters, a measure the Republican National Committee does not oppose. In March, he said Democratic efforts in Congress to expand mail-in voting would make it so “you’d never have a Republican elected in this country again”. Studies have shown neither Democrats nor Republicans benefit from a switch to a vote-by-mail system.

Trump also attacked efforts to increase mail-in voting in Michigan and Nevada last week. He falsely said Michigan was sending absentee ballots to all registered voters; in fact the state was only sending an absentee ballot application. Trump later deleted his tweet and reposted a new one accusing the secretary of state of unlawfully sending the applications, something the secretary of state says is well within her authority.

Trump’s campaign manager, Brad Parscale, lambasted Twitter’s decision in a statement, framing it as an attempt by “Silicon Valley” to “obstruct and interfere with President Trump getting his message through to voters”. Parscale also asserted that Twitter’s “clear political bias” was a reason that the Trump campaign pulled its advertising from Twitter “months ago”.

Twitter banned all political advertising in November 2019, more than six months ago. At the time, Parscale decried the global policy change as “another attempt by the left to silence Trump and conservatives”.

Related: Twitter to ban all political advertising, raising pressure on Facebook

Experts on misinformation raised questions about whether Twitter’s measure would be effective. Mike Caulfield, head of the Digital Polarization Initiative of the American Democracy Project, noted that the “Get the facts” phrasing could further entrench misinformation with its “legitimizing tone”.

“Get the facts implies there is a debate where facts are being marshalled in evidence,” he wrote on Twitter. “It elevates the claim.”

Claire Wardle, the director of First Draft News, also questioned whether Twitter’s information page would be effective in changing any minds.

“If we are to consider the reasoning behind this, it’s not a belief that this will change anyone’s minds, it’s to provide necessary context to a tweet that should be taken down, but if they did that would lead to more conspiratorial thinking,” she said. “They’re stuck between a rock and hard place.”