January 2019: #WeSupportPolandSummit
The U.S.-sponsored Warsaw Middle East Summit — which largely focused on Iran’s threatening activity in the region — was a month away (February 13th/14th, 2019), and the #WeSupportPolandSummit hashtag was trending on Twitter in Iran.
On January 15th, 2019, I noticed the following tweets from Ben Nimmo, who currently is the Global Lead for Threat Intelligence at Meta (Nimmo was employed by the Atlantic Council’s Digital Forensics Research Lab at the time, and later went on to lead investigations for CIA-backed Graphika):
Effectively, Nimmo categorizes activity in relation to the #WeSupportPolandSummit hashtag as being suspicious, concluding that it “could be bots, or troll teams trying to game the trending algorithm.”
Nimmo’s tweets put #WeSupportPolandSummit on my radar, compelling me to collect/review data of accounts using the hashtag, sharing my findings via this thread on January 17th, 2019:
After collecting more than 100,000 tweets where the #WeSupportPolandSummit hashtag appeared (inclusive of retweets; 5.5K unique tweets), it was clear that Nimmo’s assertion — namely, that the hashtag did not trend organically — was indeed correct.
In fact, Twitter’s acknowledgement that artificial amplification (i.e. fake/inauthentic accounts retweeting the hashtag) — a form of platform manipulation that violates Twitter Rules — played a role in driving the popularity of #WeSupportPolandSummit may be inferred by referencing the Google Spreadsheet linked to via my thread.
Please note that not every account in the spreadsheet is an account that is engaging in coordinated inauthentic behavior (i.e. gaming #WeSupportPolandSummit to trend); rather, the spreadsheet simply includes accounts appearing in at least 500 tweets (inclusive of retweets) where #WeSupportPolandSummit appears. Moreover, the list has nothing to do with whether an account is a bot (i.e. fully automated) or not.
The reason @realDonaldTrump is on the list, for example, is because his account appears in at least 500 tweets (inclusive of retweets) where #WeSupportPolandSummit appears. Former President Trump never tweeted/retweeted the #WeSupportPolandSummit hashtag himself, and the appearance of his (or any other) account on the list should not be misconstrued such that it equates the list with being bots.
Of the nearly 400 accounts that met said criteria, as of May 11th, 2023, 67 of them have been suspended by Twitter:
In other words, 18% of the accounts most frequently associated with #WeSupportPolandSummit in January 2019 had been suspended as of May 2023. And less than 50% of the accounts most frequently associated with the hashtag in January 2019 still exist today.
Twitter is implicitly acknowledging what Nimmo and myself have asserted — namely, that the #WeSupportPolandSummit hashtag was gamed — via their actions (i.e. account suspensions).
Over the next several weeks (January 2019 into February 2019) I would go on to document platform manipulation across the Iranian Twittersphere, extending well beyond the #WeSupportPolandSummit hashtag:
Then I noticed something odd.
My account had been blocked by @IsraelPersian, Israel’s official Persian language account:
Additionally, on January 29th, 2019, I was doxed (i.e. private information about myself and my family, including addresses & phone numbers) by a Twitter account (@no_itsmyturn) that—based on research conducted in 2021— appears to be connected to Israeli intelligence.
Shortly after realizing @IsraelPersian had blocked my account, I opted to take a quick look at the accounts being followed by Israel’s official Persian language account. It was then when I noticed @IsraelPersian was Following quite a few of the accounts that were manipulating hashtags to trend.
It wouldn’t be until several years later, however, that I would revisit the @IsraelPersian account.
@IsraelPersian/@SAvginsaz Following Map
In late 2022, the National Iranian American Council (NIAC), after experiencing an increase in online attacks directed at diplomacy-oriented Iranian-Americans, engaged my company, Social Forensics. NIAC sought to better understand the scale of the attacks, whether or not platform manipulation was involved, and if state actors were possibly contributing to the vitriol emanating across the Iranian Twittersphere.
Worth noting that prior to this point, all of the Iran-focused platform manipulation research that myself/Social Forensics have conducted has been unpaid, and driven by personal interest (i.e. receiving death threats and being doxed are powerful motivators when it comes to shining light on bad actors).
As was recently highlighted by Politico, “it is unclear who many of the assailants are, because many of the attacks are anonymous, often using social media accounts without clear provenance.”
Since I already knew that Israel’s official Persian language account (@IsraelPersian) was Following a number of accounts engaging in platform manipulation (hashtag gaming, targeted abuse), focusing on that account’s Following graph seemed to be a good place to start.
In doing so, I came across the account of Sharona Soleimany Avginsaz (@SAvginsaz) for the first time. Sharona is the Director of New Media in Persian at the Israeli Ministry of Foreign Affairs (MFA). Given her title, it seems fair to assume that Sharona is the person operating the @IsraelPersian account:
In fact, between these two official Israeli government accounts, more than 1 in 8 accounts (50 out of 396) they are Following are accounts that engage in platform manipulation:
Broadly speaking the accounts they follow fall into one of 3 categories/communities: 1) Persian Media, 2) Israeli Government, 3) Monarchist (i.e. those who advocate for the overthrow of the Islamic Republic and the return of exiled Crown Prince Reza Pahlavi to the country as the leader of a constitutional monarchy)
When an account is selected within the map, it is possible to view other accounts from the map that the selected account is Following or has as Followers.
For example, here is what accounts Reza Pahlavi is Following looks like vs. accounts that are Followers of Iran’s former Crown Prince:
Meet Hananya Naftali
Hananya Naftali (@HananyaNaftali), according to Jewish News Syndicate, is “a prominent pro-Israel influencer on multiple social media platforms, speaking out against antisemitism, BDS, Iran and other threats the Jewish people face on a daily basis.”
Since 2017, Naftali has been employed as a social media advisor to Israeli prime minister, Benjamin Netanyahu, where his focus has been public diplomacy:
When looking to Naftali’s Twitter account within the interactive map, as expected his account is Following more Israel-focused accounts (bottom of map) than Iran-focused accounts (top of map):
Moreover, among the accounts being followed is @HananyaPersian, Naftali’s Persian language account:
Tweets from Naftali, who does not speak Persian, appear quite popular to those looking at engagement (retweets, likes, replies) as a proxy for public sentiment. Take this tweet, for example, which has driven more than 400 likes just 2 hours after being posted:
In reality, however, the vast majority of the tweet’s engagement is being driven by inauthentic accounts/artificial amplification. And the same holds true for any tweet from Naftali — whether from his English (@HananyaNaftali) or Persian (@HananyaPersian) account.
Hence, I was not surprised to discover that the Persian-language account of Netanyahu’s social media advisor is Following many of the same inauthentic accounts being followed by @IsraelPersian and/or @SAvginsaz:
Professionalizing Platform Manipulation
Attempts by the IDF and other Israeli security forces to shape the public consciousness are generally kept secret. Earlier this year, however, the Israeli military admitted to “deploying keyboard warriors” as part of a covert influence campaign during the 2021 Israel–Palestine crisis. Throughout the bloody 11-day war, the Israeli military employed fake accounts — which praised their airstrikes in the Gaza Strip across Twitter, Facebook, Instagram and TikTok — in “an effort to improve the Israeli public’s view of Israel’s performance in the conflict.” Moreover, the Israeli military “conceded that it also coordinated the campaign with real social media influencers, providing them with images and hashtags to talk up the military’s achievements and showcase the damage it inflicted on Gaza.”
The Israeli Ministry of Foreign Affairs (MFA), on the other hand, has been much more open about their efforts to control the information environment when compared to the IDF and other Israeli security forces. In fact, they even opted to professionalize platform manipulation, as may be seen via reviewing LinkedIn profiles of several former Israeli MFA employees. For example, Elad Ratson — who served as an Israeli diplomat for more than a decade — spent 4 years with the Israeli MFA (2015 to 2019) and refers to his work there as “Pioneering Algorithmic Diplomacy: a code-based approach to Diplomacy.” Ratson has since moved on to the private sector where he is the founder and CEO of Vayehee. According to Ratson’s LinkedIn Profile, Vayehee provides “R&D services for digital communications, big-data harvesting, data analysis, OSINT & narrative dominance on social media.”
And then there’s Yoav Adler, who, like Ratson, has since opted to trade his Israeli MFA role for a private sector one. Adler similarly had a long-tenure (nearly a decade!) as an Israeli diplomat, rounding out his time there serving as the Director of Research & Development from May 2017 to Sep 2019. It appears that Adler replaced Ratson as the MFA’s Director of R&D; Ratson moved on to serve as Special Ministry of Foreign Affairs Adviser for Algorithmic Diplomacy and later as Head of Data Diplomacy R&D Unit.
This is how Adler’s LinkedIn profile describes his time as the MFA’s Director of Research & Development: “In my capacity as the Director of the Research and Development Unit at the Israeli Ministry of Foreign Affairs, my team and I develop innovative algorithms and software which promote and advance diplomatic objectives via online digital channels. In fact, I pioneer a new sub field of Digital Diplomacy experts refer to as Code-based or Algorithmic Diplomacy — the harnessing of algorithms to influence the flow of country related narratives in the online matrix. To the best of my knowledge this is a unique and groundbreaking position in the world of diplomacy — translating diplomatic objectives into code language.”
Effectively what Adler is describing there is computational propaganda. Computational propaganda involves the “use of algorithms, automation, and human curation to purposefully distribute misleading information over social media networks” (Woolley & Howard, 2018).
The Oxford Internet Institute does a nice job outlining computational propaganda Strategies & Targets:
OK, So What’s Your Point?
The initiative had previously been awarded a $1.5 million grant via the State Department’s Global Engagement Center (GEC) — funding that, according to the Iran Disinformation Project’s website, would be used towards bringing “to light disinformation emanating from the Islamic Republic of Iran via official rhetoric, state propaganda outlets, social media manipulation and more.”
Rather than combating foreign propaganda and disinformation (GEC’s purpose), however, the Iran Disinformation Project was using their Twitter account (@IranDisinfo) to target journalists, think tank analysts, researchers, and civil society advocates — denouncing them as “mullah sympathizers,” “apologists,” “mouthpieces,” and “lobbyists.”
It goes without saying that the U.S. government should not be funding efforts that target Americans in this fashion. Moreover, the U.S. government should be very concerned about foreign governments doing the same thing to Americans, even when those foreign governments are U.S. allies (as is the case with Israel).
In the coming weeks, Social Forensics will publicly make available a report/investigation compiled for NIAC that outlines the scale of platform manipulation across the Iranian Twittersphere, and why we believe the Israeli government is driving these efforts.
In the interim, please be sure to check out the web-based interactive @IsraelPersian/@SAvginsaz Following Map to explore on your own! https://twittermaps.socialforensics.com/israelpersian
Sunlight is the best disinfectant.
Social Forensics maps and monitors social connections and activity.
We create purposefully designed tools to manage social data analytics needs across various industries. Our focus is audience segmentation and identifying coordinated inauthentic behavior (CIB) across social media platforms.
Geoff Golberg is an NYC-based researcher (and entrepreneur) who is fascinated by graph visualization/network analysis — more specifically, when applied to social networks and blockchain activity. His experience spans structured finance, ad tech, and digital marketing/customer acquisition, both at startups and public companies.
Geoff is the Founder/CEO/Janitor of Social Forensics.