Dumping, Astroturfing and Trolling
How Intelligence Chair Richard Burr exemplifies all that is wrong with politics
Richard Burr (R) is the senior U.S. Senator from North Carolina, where he has been serving since 2005.
The Chairman of the Senate Intelligence Committee, Burr made headlines last Thursday on news that he sold a significant portion of his stock portfolio around the same time that he was receiving daily updates on the coronavirus health threat (and prior to the markets tanking).
Despite raising alarm (re: coronavirus preparedness) privately at a luncheon last month (where a small group of well-connected constituents were in attendance), Burr took a far less dire tone when addressing the public at large.
Cambridge Analytica
Cambridge Analytica broke into the U.S. political scene in 2013, testing psychographic messaging in Virginia’s gubernatorial race (and in support of Republican Ken Cuccinelli). Despite Cuccinelli losing the election, the political data strategy appeared promising, resulting in Steve Bannon partnering with Alexander Nix (SCL’s elections director at the time) to create Cambridge Analytica.
During the 2014 election cycle, North Carolina senator Thom Tillis (R), North Carolina representative Patrick McHenry (R), and the North Carolina GOP used Cambridge Analytica — collectively spending $360K on various media, research, data management, and voter/donor modeling services.
Senator Burr contracted with Cambridge Analytica during the 2016 election cycle, benefiting from survey and other research both directly (his campaign’s contract with CA) and indirectly (via John Bolton Super PAC’s contract with CA, $811K; Burr’s campaign received $338K in support from the super PAC during 2016 cycle).
Cambridge Analytica, of course, is best known for their work on President Trump’s campaign. The political data analytics firm came under scrutiny in 2018 for their business practices (more specifically, their approach to harvesting Facebook data). Facebook suspended Cambridge Analytica in March 2018, citing misuse of user data. On May 1st, 2018, Cambridge Analytica and its parent company filed for insolvency proceedings and closed operations.
Twitter Shenanigans
Two years ago, I discovered (and tweeted about) a family of Twitter accounts that were falsely representing North Carolina and various North Carolina municipalities:
The deceptive Greensboro account would frequently tweet localized content, including lots of images and regularly adding geotags to their tweets (functioning to increase discoverability). Often the account would post from parks, museums, and sporting events (among other venues!). Greensboro residents with an interest in the arts could conceivably discover that a seasonal exhibit was recently installed in a local museum, after encountering a tweet from an account they were Following — namely, an account they believed to be the official Twitter account of Greensboro.
Suffice it to say that while the vast majority of the account’s ~256K Followers were inauthentic accounts, the impostor Greensboro account’s Followers also included many (real) North Carolina residents.
Similar audience-building tactics were employed by other accounts from the group.
After conditioning their Followers to expect localized content, many of these accounts would inject (retweet) Trump/right-leaning content in a coordinated fashion into the feeds of their respective Followers (artificial amplification violates Twitter Rules).
Imagine the sort of havoc an account like @greensboro_nc could wreak during election time — tweeting “polls are closing early today”, for example.
Between March and May (2018), I was in touch with several Twitter execs/senior employees via Twitter DM and email; additionally, I would frequently tag them in tweets highlighting my concern around the North Carolina accounts. I know my message was heard (loud and clear!), as most DMs/emails were actually answered. Still, I wasn’t taken seriously.
Mike Farb similarly felt this was an important issue for Twitter to address and followed up with this fantastic thread in June 2018:
Mike’s thread, fortunately, resulted in coverage by The News & Observer (based in Raleigh; serves the greater NC Triangle area). A few days later, Twitter (finally) removed the main accounts. Worth noting, however, that Twitter did not remove hundreds of thousands of fake/bot accounts which were following said main accounts.
To be clear, the issue with the family of North Carolina accounts has nothing to do with them amplifying Trump/right-leaning content. In fact, empowering users/accounts to amplify content is, in part, why Twitter’s platform is so powerful/special.
The problem is many of those accounts appeared to be far more influential than they actually were, as many of their Followers were comprised of fake/bot/inactive accounts. The intent behind the operation was to deceive, and the accounts should have been suspended much earlier for being in violation of Twitter Rules.
Hence, Twitter’s (better late than never) suspension of the accounts was driven by account behavior, rather than the actual content being shared.
When Shit Got Weird
Scanning the Followers of @_North_Carolina (falsely representing North Carolina) and @greensboro_nc (falsely representing Greensboro) accounts, I was taken aback when discovering the accounts of both North Carolina senators (Thom Tillis and Richard Burr):
Were Senators Tillis and Burr duped? Or, were they Following the imposter accounts because they were somehow involved?
Next, I checked if the pair of North Carolina senators were Following the real Greensboro account (@greensborocity).
I knew something funky was going on immediately after seeing the senators were not Following the official account of Greensboro (North Carolina’s third largest municipality) — and in light of the fact that the senators were following the deceptive Greensboro account:
Meet @BurrForSenate
Senator Burr maintains (at least) two Twitter accounts (@SenatorBurr and @BurrForSenate).
Analysis to follow will focus on @BurrForSenate, an account used through Senator Burr’s 2016 re-election (the account hasn’t tweeted or liked any tweets since November 2016):
Of the 6,528 accounts being followed by @BurrForSenate (i.e. Following), 74% of accounts (4,800) are also following @BurrForSenate (i.e. Followers).
Of the 8,512 accounts following @BurrForSenate (i.e. Followers), 56% of accounts (4,800) are also being followed by @BurrForSenate (i.e. Following).
The 654th account to follow @BurrForSenate was @GOPAlert2017 (which goes by the display name “DJT ON A KILLSTREAK”):
The @GOPAlert2017 account was created in May 2015 (May 14th, 2015):
Translated: Followers 654 through 8,512 (7,859 accounts) of @BurrForSenate took place on or after May 14th, 2015 (the creation date of @GOPAlert2017)
The @BurrForSenate account was created in November 2010 (November 8th, 2010). Up until the time the account was followed by @GOPAlert2017, it had only tweeted 3 times in 4.5 years. Those tweets are below:
In June 2015 — and just under a month after @GOPAlert2017 was created — the account of Senator Burr’s campaign committee announced they were “back” on Twitter:
Note: the @GOPAlert2017 was selected merely to better understand timeline (i.e. @BurrForSenate picked up 7,859 of their 8,112 Followers on or after May 14th, 2015, despite the account being created in November 2010)
Here’s what the timing of overall tweets from @BurrForSenate looks like (once again, the account hasn’t tweeted or liked any tweets since November 2016):
In summary, the vast majority (93%) of tweets from @BurrForSenate took place during 2016 and the account hasn’t tweeted since November 2016:
Translated: @BurrForSenate has somehow managed to gain nearly 8 thousand Followers after spending a mere 1.5 years actively tweeting from the account (104 tweets in 2015; 1,370 tweets in 2016)
Follow-For-Follow
Given the @BurrForSenate account shares reciprocal following relationships (i.e. accounts that appear in both Following and Followers of @BurrForSenate) with 4,800 accounts — and after only 18 months of activity — it would be fair to categorize @BurrForSenate as a follow-for-follow account.
As was highlighted earlier in the post, the 4,800 accounts represent 74% of @BurrForSenate’s Following (6,528 accounts) and 56% of @BurrForSenate’s Followers (8,512 accounts).
More specifically, follow-for-follow accounts are generally Following a similar number of accounts as they have as Followers. Moreover, follow-for-follow accounts will be Following a large percentage of the accounts that comprise their Followers.
We can get a sense of the follow-for-follow nature by comparing the overall orders (Following and Followers) of the 4,800 accounts that are both following and being followed by @BurrForSenate (below sorted by Fllwing order, earliest to most recent; 30 account snapshot):
For starters, the official Twitter account of Intelligence Chair Burr’s campaign committee is (reciprocally) Following “Wash your hands, you animal!” (who happens to be a #CommonSense #ProMilitary type using the handle @obligatoryasian):
Here’s a summary of Followers counts for the 8,512 accounts that follow @BurrForSenate:
There are 39 unverified accounts, for example, that follow @BurrForSenate which have between 200,000 and 499,999 Followers. Meanwhile, there are 1,289 unverified accounts that follow @BurrForSenate which have between 5,000 and 39,999 Followers.
The 39 unverified accounts (with Followers between 200,000 and 499,999), on average, reflect having Following counts that are the same as Followers counts (avg ratio = 1.01).
The 1,289 unverified accounts (with Followers between 5,000 and 39,999), on average, reflect having Following counts that are 94% of Followers counts (ratio = Following/Followers; avg ratio = .94).
It is worth noting that Twitter Rules prohibit “aggressive following”:
Whereas on the one hand there are nearly 1,600 unverified (mainly follow-for-follow) accounts (with at least 5,000 Followers) that are following @BurrForSenate, on the other hand, there are also a similar number of unverified accounts (with less than 100 Followers) that are following the official Twitter account of Intelligence Chair Burr’s campaign committee.
Ugh, Senator.. Why Are You Following THOSE Accounts?
Visuals are often more powerful than words.
Here are the 30 accounts from the table presented earlier (where @obligatoryasian appeared; each of the 30 accounts are both following and being followed by @BurrForSenate):
Intelligence Chair Richard Burr, like a number of other Republican politicians, seems to have an affinity for QAnon — the @BurrForSenate account is Following 17 accounts that display their affection for the conspiracy theory in their account bios:
Additionally, @BurrForSenate is Following 58 accounts that include WWG1WGA (“where we go one, we go all”), a slogan/rallying cry adopted by QAnon supporters, in their account bios:
In aggregate, accounts being followed by @BurrForSenate that include either QAnon or WWG1GA in their bios, have generated 5.2 million tweets.
Creating The Illusion of Support
Astroturfing, defined below by Wikipedia, is a common tactic utilized in the political realm (and where social platforms often function as conduits):
In the case of the impostor @greensboro_nc account, for example, Greensboro municipal staffers were dismissed by Twitter when trying to get the Big Tech company to remove the deceptive account:
Referencing the tweets being amplified by the impostor account (i.e. supporting conservative/right-wing issues), Jake Keys, the municipal staffer who has run Greensboro’s official Twitter account for several years, stated “I don’t want people to think that’s what the city of Greensboro municipal account would behave like.. I would really love to have that [account] gone or dismissed. It can be confusing.”
Those behind this North Carolina-focused information operations effort, on the other hand, were (and continue to be) wildly successful. The imposter Greensboro account alone had churned out more than a million tweets prior to media coverage (finally) forcing Twitter’s hand to act. Said operators were able to build Twitter followings that included actual North Carolina citizens — establishing valuable distribution infrastructure, in the process, that continues to allow for (coordinated, inauthentic) attempts at manipulating North Carolina voters.
Specific to @BurrForSenate, of the account’s 8,512 Followers, just over three quarters (6,459 accounts) include a non-blank entry in the location field of their bios. Of the accounts with non-blank location fields, at least 36% of accounts (2,334) indicate that a Twitter user located in North Carolina is the account owner (important to keep in mind that users input the location field vs. being detected by Twitter).
Here are examples of several (11) of the more active accounts (each having tweeted at least 100 thousand times) that are reciprocally following @BurrForSenate (and purporting to be located in North Carolina):
It’s certainly odd that the official Twitter account of Intelligence Chair Burr’s campaign committee is Following so many accounts that violate Twitter Rules.
By representing themselves as being located in North Carolina, these accounts seek to create the illusion of support—namely, that support for conservative/right-wing issues is widespread among residents of North Carolina.
Trolling
The Chairman of the Senate Intelligence Committee is a troll.
Senator Burr’s recent warning to his constituents (re: coronavirus preparedness) during a private luncheon isn’t the first time Burr’s remarks from a private event made their way to the press via leaked recordings.
In October 2016—and in the midst of a tight reelection race against Deborah Ross — Burr, while addressing GOP volunteers in Mooresville, North Carolina, “privately mused.. [that].. gun owners may want to put a “bullseye” on Hillary Clinton, according to audio obtained by CNN.”
Around the same time that Senator Burr himself was trolling Hilary Clinton in person (privately among his supporters), the official Twitter account of Intelligence Chair Burr’s campaign committee was publicly trolling Deborah Ross via an aggressive smear campaign, heavily amplified (retweeted) by inauthentic accounts:
Taking a closer look into the accounts retweeting the above tweet, well.. as was mentioned earlier, sometimes it’s best simply to leave it to the visuals:
Here are a few more tweets from @BurrForSenate in September/October 2018 (and similarly targeting Deborah Ross):
Note how the Twitter account of Deborah Ross (@DeborahRossNC) was included in the (4) tweets above. This functions to jam/spam the Twitter notifications of Mrs. Ross — every time there is a retweet, like, or reply to the tweets from @BurrForSenate it makes it more difficult for her (and her campaign) to effectively utilize Twitter (and at a critical time in the race, no less).
Also worth calling attention to the hashtags being used. The tweets directed at Ross (4 directly above and the single tweet above those), for example, each include #ncsen, while 2 (of the 5) include #ncpol as well.
The hashtag use coupled with strong engagement (again, being driven by inauthentic accounts) ensures said tweets appear towards the top of Twitter searches for the hashtags.
OK, So What’s Your Point?
Burr led the Senate’s investigation into whether the Trump campaign colluded with Russia to interfere in 2016 U.S. presidential elections. Considering both Burr and Trump campaigns were contracting with Cambridge Analytica at that time, this is quite remarkable.
Although Cambridge Analytica focused on Facebook data (i.e. it had nothing to do with Twitter), it shouldn’t come as much of a surprise that campaigns leveraging data obtained via questionable means, would seek to extend their information operations efforts beyond Facebook’s platform.
To be clear, I am not stating that Senator Burr himself (someone from his staff is far more likely) is behind his official campaign committee’s Twitter account Following accounts like “TriggaNometry” (@grindingdude; account has more than 86K Followers; IFB, included in the bio, stands for “I follow back”):
Currently, social media platforms are struggling (big time, in fact) at effectively mitigating coronavirus-related information pollution issues.
Starting in 2018 (and in response heavy criticism), Facebook and Twitter began rolling out new features (and policies) aimed at ensuring their platforms aren’t weaponized by nefarious actors seeking to sway elections (i.e. avoiding another 2016 U.S. presidential elections scenario).
Building out new features and modifying platform rules/policies, however, is rendered useless when platforms fail to consistently (and in a non-selective fashion) apply said rules.
Unfortunately, that continues to be the case for Twitter—who, after recently issuing new policy focused on taking a harsh stance against coronavirus-related mis/disinformation, simply isn’t enforcing it.
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, where he spends (far too much of) his time developing techniques and building tools to identify social media manipulation (of various flavors!).
Geoff’s Twitter account was mass reported by nefarious actors who sought to silence his voice, resulting in the automated suspension of his Twitter account in July 2019.
Geoff recently (March 3rd, 2020) filed a lawsuit against Twitter for banning his Twitter account in bad faith — and he strives to prevent Twitter from unlawfully cancelling the accounts of other New York residents in the absence of cause.
