Twitter Inc. has removed more than 10,000 automated accounts that reported sharing in which people were discouraged from voting in the upcoming US Midterm elections.
The US Democratic party contacted Twitter after it appeared that the fake accounts were showing that they were Democrats. For example, some tweets called men who wanted to vote for the Democrats not to go to the ballot box because their voices would oppress women.
“We took action against the accounts and activity on Twitter,” a company spokesperson told Reuters by e-mail. The measures date from the end of September and the beginning of October. According to three sources that have links with the Democratic party, there are more than 10,000 fake accounts. The Campaign Committee of the Democratic Party (DCCC) developed its own system to detect such automated accounts on social media and works together with various companies. These efforts came in response to the millions of accounts on Twitter and other social media that spread negative and false information during the 2016 US presidential election about the then Democratic candidate Hillary Clinton and other party members.
Who is responsible for these fake accounts is unclear. “We cannot say we are behind these operations. Twitter keeps it hidden from us, but with this technology you know when and how it happens,” says Ash Bhat, co-founder of RoBhat Labs, one of the companies that detect bots for the DCCC and identify political bias in messages.