Researchers Find No Proof of Covert Social Media Network Working to Boost Netanyahu

The Big Bots Project had alleged that the network included 154 accounts using fake names, with another 400 accounts suspected of being fake

Supporters of the Likud Party hold a photo of Prime Minister Benjamin Netanyahu at the launch of Likud party's election campaign in Ramat Gan, Israel March 4, 2019.
REUTERS/Amir Cohen

Following the release of a watchdog group’s report last week alleging that a network of hundreds of social media accounts has been working to boost Prime Minister Benjamin Netanyahu’s re-election campaign, researchers at Ben-Gurion University of the Negev who examined the report say that the data presented does not prove the existence of a coordinated operation meant to influence the election.

Instead, the Ben-Gurion researchers claim, the report reflects a pattern that is typical of legitimate political activity on social media.

The researchers, Dr. Oren Tsur, and two graduate students, from the university’s natural language processing and social dynamics lab, have concluded that “the original report suffers from a number of methodological problems that make it impossible to draw statistically significant conclusions” regarding the alleged unusual timing of social media posts purportedly designed to manipulate the election.

Haaretz Weekly Ep. 21Haaretz

>> Phone hacking? Fake Twitter accounts? You ain't seen nothing yet | Analysis ■ Netanyahu presides over a social media empire. Here's how he runs it

The report from the watchdog group, the Big Bots Project, was initially reported by the Israeli daily Yedioth Aharonoth and the New York Times. The issue was taken up last week at a hearing by the Central Election Committee, chaired by Supreme Court Justice Hanan Melcer. The Ben-Gurion researchers said they sought to answer a question posed at the hearing by Justice Melcer regarding whether the social media activity in support of Netanyahu could be explained by reasons other than a coordinated effort.

Tsur told Haaretz that by its very nature, social media outlets are coordinated in one manner or another without the need for a command structure or paid staff sending out messages.

The researchers don’t discount what remains the strongest evidence of a coordinated effort by paid staff – comments by Yitzhak Haddad, who purportedly told a private investigator that he was a paid staffer (a claim that was later denied by his lawyer, who threatened to file a defamation suit). Tsur also didn’t exclude the possibility that there could be networks working on behalf of one or another political party or that there was coordination behind the scenes, but he said the data on social media activity presented in the report do not prove that the social media accounts were used in an illegitimate manner.

The Ben-Gurion researchers welcomed the existence of the Big Bots Project, but noted that it was run primarily by volunteers. Tsur and his colleagues were also highly critical of how it was covered by the news media, saying that the study’s methodology was not adequately examined.

While the study found no direct links between the network and Netanyahu or his Likud party, the Big Bots Project said the accounts appeared to work in tandem with one another and in coordination with Likud’s election campaign. The network included 154 accounts using fake names, with another 400 accounts suspected of being fake, with all of them apparently being operated by real people and with the posts receiving over 2.5 million hits, the report said.

All the accounts were linked to Haddad of Ashdod, according to the report, which cited a YouTube channel to which he is an active subscriber that features a message offering money for “responding on Facebook and on the internet with political messages. You just get political messages and you post them.” 

However, the researchers point out that not all social media coordination is improper.

“The very essence of social media is coordination that develops organically,” the Ben-Gurion researchers stated. “We have shown that such coordination clearly statistically exists among various groups from all ends of the political spectrum. It’s a statistical-social phenomenon that results from the nature of social media and from the major interest in dramatic events and does not involve evidence of improper coordination.”