Twitter Execs Are Lying To Us
How Twitter’s negligence results in nefarious actors distorting the public debate
Earlier this month, I took a look at #TrudeauCorruption, highlighting how inauthentic accounts aided the hashtag in trending.
Later that day, Marc Owen Jones, an assistant professor of Middle East studies and digital humanities at Hamad bin Khalifa University, posted a fantastic thread sharing his findings related to the #TrudeauMustGo and #TrudeauCorruption hashtags:
Marc’s conclusion — namely, that it “seems people’s instinctive suspicions that the hashtag is manipulated by an organized disinformation campaign are valid” — is absolutely correct.
Back in July, National Observer’s Caroline Orr delved into #TrudeauMustGo — highlighting that “much of the activity surrounding the hashtag was actually driven by accounts tweeting at non-human rates, including about two dozen accounts created in the past 48 hours.”
Yet, despite multiple researchers independently highlighting what appears to be blatant platform manipulation, Twitter insists on singing a different tune.
Last week, Twitter Canada’s head of government and public policy, Michele Austin, told CBC/Radio-Canada that Twitter’s “initial investigation has not found evidence of bot activity amplifying the #TrudeauMustGo hashtag.” Moreover, Mrs. Austin even doubles down on her confidence, asserting that “these were driven by organic, authentic conversation.”
Organic, Authentic Conversation
Auntie Trudeau
One of the accounts highlighted in my post was “Auntie Trudeau” (@AuntieTrudeau) — who, according to their Twitter bio, believes Canada needs to “elect @CPC_HQ [Conservative Party”] to stop [their] decent (sic) into a socialist abyss”:
“Auntie Trudeau” has, without a doubt, contributed to the amplification of several anti-Trudeau hashtags. For example, between July 25th and September 5th (42 days), @AuntieTrudeau tweeted #TrudeauMustGo 839 times (on average, 20 times per day).
Is this what ‘organic, authentic conversation’ looks like?
Clearly the answer is no, as Twitter suspended the account of “Auntie Trudeau” on September 9th (and a few days after my post):
As reflected in the graph below, @AuntieTrudeau was one of the more central/influential accounts that contributed to #TrudeauCorruption earning valuable real estate across Twitter’s Trending topics:
Joe: The CBC special libtard dish, very spicy indeed.
Another account highlighted in my post was “Just an average Joe from Canada…” (@Working_Cdn2019):
Back when I wrote that post, the account had spent the prior 9 days, on average, tweeting more than 350 times each day. Between August 27th and September 5th (9 days), @Working_Cdn2019 tweeted #TrudeauMustGo 1,560 times (on average, more than 170 times per day!). Additionally, “Joe” had tweeted 178K since January 2013.
The account has, without a doubt, contributed to the amplification of the hashtag — one that has, on several occasions, made its way to Twitter’s Trending topics.
Is this what ‘organic, authentic conversation’ looks like? (Note how “Joe” would frequently amplify “Auntie Trudeau”)
“Just an average Joe from Canada…” is no more; instead, the account’s operator has opted to change its name to “The CBC special libtard dish, very spicy indeed.” (@ThePoutineGate; userid=1093277930):
Nefarious actors frequently change their Twitter usernames. This is done to evade Twitter’s detection (while also making it more difficult for the average user to keep track of their accounts).
As reflected in the graph below, “The CBC special libtard dish, very spicy indeed.” (fka “Just an average Joe from Canada…”) was one of the more central/influential accounts that contributed to #TrudeauCorruption earning valuable real estate across Twitter’s Trending topics (@Working_Cdn2019; highlighted in the image on the right):
Earlier today (September 21st, 2019), Twitter finally took action, suspending the @ThePoutineGate account:
As a reminder, Twitter Canada’s head of government and public policy, Michele Austin, has asserted that these hashtags “were driven by organic, authentic conversation.”
If the hashtags are being ‘driven by organic, authentic conversation,’ then why is Twitter suspending accounts contributing to the amplification of said hashtags?
Tip Of The Iceberg
Twitter’s reactive approach to mitigating against platform manipulation/information operations can generally be characterized by superficial efforts — doing the bare minimum to get media off their backs, basically.
I can state with certainty that Twitter Canada — despite what Mrs. Austin has communicated — currently has a major problem with nefarious actors distorting the public debate. Hashtags are regularly gamed to trend via artificial amplification, politicians/journalists are being spammed/swarmed by inauthentic accounts, and trolls like “Auntie Trudeau” and “Joe” continue to abuse Canadian citizens.
Below I have highlighted a subset of hyperactive accounts (31 accounts; each has averaged 100+ tweets per day since being created) from the #TrudeauCorruption dataset:
Of the 31 accounts, Twitter has only acted on one (suspending the @UpKeeks account).
By itself, tweeting 100+ times per day doesn’t necessarily mean an account is engaging in nefarious activity. Many news/media organizations, for example, leverage automation to tweet hundreds of times per day — hoping to reach different audiences throughout the day with their content.
After reviewing each of the 31 accounts in greater detail, however, I can state with certainty that all of them should have been suspended long ago for violating Twitter Rules. NOTE: I am not stating that ANY of the accounts are “bots” (moreover, the word “bot” is not used in this post aside from this section and when referencing a direct statement from Twitter’s Michele Austin)
Here is an overview of each account’s profile data:
Accountability
Jack Dorsey’s pinned tweet (from March 2018) states: “We’re committing Twitter to help increase the collective health, openness, and civility of public conversation, and to hold ourselves publicly accountable towards progress”:
Vijaya Gadde, Twitter’s Legal, Policy and Trust & Safety Lead tweeted the following a couple week’s prior to Jack’s call for public accountability:
Vijaya’s claim that Twitter ‘proactively [identifies] suspicious account behaviors’ simply doesn’t align with what I have seen (and have been highlighting since early 2018).
Twitter executives are lying to us.
Twitter leadership is lying to us (as this post highlights re: Mrs. Austin’s statements).
If it’s public accountability that Jack Dorsey is looking for, well, then, Jack, please stop silencing researchers like myself for attempting to hold you publicly accountable.
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!).
Read about Geoff’s war with Twitter here!
