One of the most popular accounts, called “Wake Up America,” shared images that criticized Trump, calling him the “worst, and most hated president.” Overall, the page had more than 53,000 “likes,” according to data released by the Atlantic Council’s Digital Forensic Research (DFR) Lab. Other removed accounts posted about Supreme Court Justice Brett M. Kavanaugh and issues such as race and immigration. They tended to echo left-leaning positions on political issues.
Nathaniel Gleicher, the head of cybersecurity policy at Facebook, said the company could not “assess the motivations of these bad actors.” But he described it as an example of “smart, well-funded adversaries who will never give up and constantly change tactics.”
Facebook’s announcement on Friday — less than two weeks before the 2018 election — illustrates that malicious actors, potentially including foreign governments, continue to use social media to push their political narratives and sow social unrest online.
Better technology and the hiring of thousands of new employees have helped social media sites more quickly find and take down propaganda and other problematic content, along with a flood of fake accounts. But Facebook and its peers face a scourge of disinformation that has become more global and sophisticated in the years since they discovered Russian operatives stirred tensions online in the run-up to the 2016 presidential election, experts say.
“The Russian playbook is out in the open, and more bad actors are going to take advantage of it,” said Sen. Mark R. Warner (Va.), the top Democrat on the Senate Intelligence Committee. “Social media companies are going to have to be more proactive, and Congress is going to have to step up and enact some guardrails so that disinformation and misinformation aren’t able to flourish so readily on these platforms.”
Facebook said Friday that it has shared the information with its counterparts, including Google and Twitter. Twitter said it had removed a small number of accounts based on information Facebook supplied. A spokeswoman for Google did not immediately respond to a request for comment.
Researchers studying Facebook’s latest revelations said the Iranian accounts demonstrated a new sophistication. They injected themselves into mainstream political debates and shared similar memes across multiple pages, much like Russia’s Internet Research Agency, which was the leading source of foreign disinformation during the 2016 election.
“The big picture is this looks like an Iranian operation which has learned from the Russian operations,” said Ben Nimmo, the information defense fellow with the DFR Lab.
Several examples provided by Facebook, however, show that the Iranian effort was marked by inconsistencies, including mistakes in grammar and punctuation. A post on the popular “Wake Up America” about a bribery allegation against a Trump associate, for example, had the dollar sign on the wrong side of the number, calling it a “secret 10$ bribe.”
Another post, featuring former San Francisco 49ers quarterback Colin Kapernick, who has emerged as a cultural flash point by leading a movement to kneel during the national anthem before National Football League games, starts with the ungrammatical phrase, “You mad at me for kneeling.” Later the same post, by a group called “No racism no war,” referred to “Indian American Genocide,” instead of the standard term, “American Indian.”
“The Russian stuff we’ve seen is really blending well with the audience they are targeting,” said Camille Francois, research and analysis director for network analysis firm Graphika, which has studied previous releases of Iranian and Russian disinformation. “The stuff that I’ve seen from Iran is a little bit less specific in the messaging and is casting a wider net.”
The takedowns are Facebook’s second action targeting accounts with ties to Iran. In August, the social networking giant removed hundreds of profiles and pages on the site and on Instagram. Facebook more explicitly tied some of those accounts to Iranian state media, as they had posted content, created events and purchased ads that at times pushed political messages sympathetic to the Iranian government.
Twitter, which also closed some Iranian accounts in August, last week released 1 million tweets from those accounts. Researchers studying the trove have found that tweets targeting U.S. audiences frequently were critical of President Trump, who has harshly criticized Iran and the nuclear deal that country negotiated with then-President Obama in 2015.
“They’re consistently talking about Trump,” said social media researcher Jeremy Blackburn, an assistant professor of computer science at the University of Alabama at Birmingham.