Research links pro-Trump, anti-Macron Twitter bots

Research links pro-Trump, anti-Macron Twitter bots
© Getty Images

A computer scientist says there are links between Twitter bots that circulated pro-Trump messages ahead of the 2016 election and bots that engaged in a disinformation campaign against French President Emmanuel Macron while he was a candidate.

Emilio Ferrara, a computer scientist at the University of Southern California, made the discovery when analyzing the Twitter campaign organized ahead of the May French election that involved spreading negative information about Macron connected to a leak of hacked campaign data. 

ADVERTISEMENT

Ferrara said many of the bots involved in attacking Macron had been created at the beginning of November — just before the U.S. presidential election — and went dormant after the U.S. election.

Months later, they again began communicating, this time spreading positive messages about French National Front party leader Marine Le Pen or negative information about Macron. 

“Many of the bot accounts that we had identified … were created in early November and active during that week,” Ferrara said in an interview. “Basically after Nov. 8, you don’t see those accounts post anything at all until the beginning of May.” 

The findings point to the “possible existence of a black-market for reusable political-disinformation bots,” the research concludes. 

Through an analysis of nearly 17 million Twitter posts made between April 27 and May 7, Ferrara and his team turned up roughly 100,000 Twitter users engaged in conversation about the leaked emails from Macron’s campaign, about 18,000 of which they determined to be bots — social media accounts controlled by computer scripts that masquerade as human users.

A large majority of the users engaged in the conversation, he said, had a history of showing support for Trump, Republican or far-right narratives. Many of them were also English speakers, he said. 

The scale of the bot operation in the French election, Ferrara observed, was much smaller than that related to the 2016 presidential election. He detected between 400,000 and half a million Twitter bots engaged in the U.S. political discussion, most of them pushing pro-Trump or far-right narratives. A minority of the suspected bots supported Democrat Hillary ClintonHillary Diane Rodham ClintonRoger Stone fundraising off promise not to testify against Trump Rivaling chants of 'USA,' 'lock him up' greet Flynn after sentencing hearing The Hill's 12:30 Report — Flynn awaits sentencing | White House signals it wants to avoid shutdown MORE

Ferrara submitted his research on the “Macron Leaks” disinformation effort, which was first reported by Motherboard, for peer review this week. 

The leak of Macron’s campaign data took place days before the French presidential election and was used by right-wing personalities and social media bots to push unsubstantiated allegations of tax fraud.

The U.S. intelligence community has concluded that Russia was behind the cyber and disinformation campaign against the American presidential election.

The French government’s cybersecurity agency investigated the Macron hack and found no evidence that a Russian hacker group tied to the U.S. presidential election was involved, the agency’s head told The Associated Press last month.

However, researchers at cyber firm Trend Micro said in April that hackers behind the hack of the Democratic National Committee had also targeted Macron’s campaign.

Ferrara said he has no evidence that Russia was behind the bot operation against Macron, noting the difficulty of making such an attribution. 

“We don’t have evidence to point to the actors,” Ferrara said. 

Experts have previously observed Russian efforts to use fake news and social media trolls to disrupt elections in Europe, including the French election, following the U.S. disinformation campaign. Germany is set to hold its federal election at the end of September.