(Bloomberg) -- Twitter Inc. removed malicious accounts originating from Iran, Russia and Venezuela that tweeted about U.S. midterm elections, the latest attempt by the social-media company to fight bad actors on its service.
The company said it continues to see activity on the service related to the Russian Internet Research Agency, the troll farm that spread divisive information during the U.S. Presidential Election. It found 418 additional accounts that appeared to originate in Russia and posted 73,398 tweets related to the midterm elections with hashtags like #MAGA and #ReleasetheMemo.
Last September, Twitter disclosed that it suspended 770 accounts potentially based in Iran for violating its policies. Since then, the company has suspended 2,617 additional malicious accounts that it believes had origins in Iran. Those accounts tweeted 24,000 times during the midterm elections, representing less than 1 percent of all Tweets from those accounts. Twitter suspended 764 accounts in Venezuela before the midterm election day and notified law enforcement of the activity.
"In contrast to 2016, we identified much less platform manipulation from bad-faith actors located abroad," Carlos Monje, Twitter’s head of Public Policy wrote in a blog post. "That said, as part of our ongoing review we found limited operations that have the potential to be connected to sources within Iran, Venezuela, and Russia."
Since the revelations around foreign propaganda and manipulation during the 2019 U.S. Presidential elections, Twitter and its social media peers like Facebook Inc. have been taking aggressive measures to fight abuse. Since January 2017, Twitter said it’s rolled out around 70 product changes and experiments, changed dozens of policies, and expanded its enforcement operations.
The 2018 U.S. midterm elections were the most Tweeted-about midterm election in history, with more than 99 million Tweets sent from the first primaries in March through Election day, according to the company’s review. Twitter took action on nearly 6,000 tweets that were deterring eligible voters, often through voter intimidation or providing false information about voting. Twitter attached an example of a tweet that read "text ’BLUEWAVE’ to 55463 on election day and VOTE from the COMFORT of your home" with logos of news organizations at the bottom of the image.
The company is now removing 214 percent more accounts year-over-year for violating its policies. For instance, over the past few months, Twitter identified and challenged as many as 10 million accounts each week suspected of spam or misusing automation.
To contact the reporter on this story: Selina Wang in San Francisco at firstname.lastname@example.org
To contact the editors responsible for this story: Jillian Ward at email@example.com, Alistair Barr
For more articles like this, please visit us at bloomberg.com
©2019 Bloomberg L.P.