NEW YORK – Under pressure to deter misuse of its platform, Twitter said Wednesday that it's adding special labels to tweets from some U.S. political candidates ahead of this year's midterm elections.
Twitter said the move is to provide users with "authentic information" and prevent spoofed and fake accounts from fooling users. The labels will include what office a person is running for and the district number of the race. The labels will sport a small icon of a government building.
Both the candidate's account and the tweets from it will get labels. The labels will appear on retweets as well as tweets off of Twitter, such as when they are embedded in a news story.
"Providing the public with authentic, trustworthy information is crucial to the democratic process, and we are committed to furthering that goal through the tools we continue to build," Twitter said in a post announcing the policy.
The labels will start to appear next week for candidates for governor and Congress. Twitter wouldn't say whether it will extend this outside the U.S., where elections interference has also been a big issue.
The San Francisco-based company is working with the nonprofit, nonpartisan Ballotpedia to help identify the relevant Twitter accounts.
Twitter, along with Facebook and other social media companies, has been under heavy scrutiny for allowing their platforms to be misused by malicious actors trying to influence elections around the world. Twitter last year disclosed that it had found roughly 200 accounts linked to the same Russian accounts that published material seemingly aimed at influencing the 2016 presidential election.
Beyond labelling accounts, Twitter and other companies have previously said they would label political ads as such and provide information on who paid for them.
Facebook said in March that it is making progress in addressing election abuse ahead of the U.S. midterms. Its efforts include expanding its fact-checking efforts and using artificial intelligence to block malicious accounts before they can spread misinformation.
for more features.