Twitter’s own disclosures identified thousands of Internet Research Agency accounts pushing Brexit content, while RT and Sputnik out-reached the official Leave campaigns on social media.
In the run-up to the United Kingdom’s 23 June 2016 referendum on European Union membership, Russian state broadcasters and covert social media accounts linked to the St. Petersburg-based Internet Research Agency (IRA) — the same troll factory later indicted by US prosecutors for interference in the 2016 US presidential election — pushed pro-Leave, anti-EU content at scale. This combined two distinct tactics: coordinated inauthentic behavior — networks of fake or hidden accounts acting in concert to make a fringe view look like widespread organic opinion — on social platforms, and astroturfing — manufacturing the appearance of grassroots public sentiment — through amplification of hashtags and talking points. Twitter’s 2018 disclosure identified thousands of IRA-linked accounts that collectively posted over a million tweets referencing Brexit, including a concentrated burst of activity on referendum day itself. Academic researchers separately mapped a broader network of roughly 150,000 Russia-linked accounts discussing Brexit before the vote. A 2020 UK parliamentary inquiry found RT and Sputnik reached more Twitter users with anti-EU content than the official Vote Leave and Leave.EU campaigns combined, while also concluding the government had never formally investigated the referendum’s vulnerability to interference.
The Twitter Disclosure
In late 2018, Twitter released a dataset of accounts it had identified as linked to the Internet Research Agency, as part of transparency efforts following the platform’s role in the 2016 US election. Researchers at the University of Edinburgh and elsewhere mined this “Brexit Twitter archive,” finding several thousand IRA-linked accounts had posted content referencing the UK, the EU, and migration in the weeks around the referendum. One widely cited count identified 2,752 IRA-linked accounts, of which 419 posted roughly 3,485 tweets specifically discussing Brexit-related topics. Separately, Twitter’s own figures cited in press coverage put the combined IRA and Iran-linked total at 3,841 and 770 accounts respectively, sending over 10 million tweets in total campaign activity that included a “day-long blitz” of posting on the day of the referendum itself.
Scale: From Thousands to 150,000 Accounts
The IRA-attributed accounts represent only the confirmed, platform-disclosed layer of activity. A separate study by researchers at Swansea University and the University of California, Berkeley identified approximately 150,000 accounts with Russian links that tweeted about Brexit in the run-up to the vote — a figure roughly forty times larger than the confirmed IRA set. A National Bureau of Economic Research working paper went further, estimating that automated bot accounts may have shifted the final Leave vote share by as much as 1.76 percentage points — a margin that, if accurate, would be larger than several of the closest recent UK general election swings, though the estimate remains contested among researchers.
State Media’s Advantage: RT and Sputnik’s Reach
Alongside covert account activity, Russia’s overt state broadcasters RT and Sputnik provided consistently one-sided coverage of the referendum campaign, a framing later confirmed by a US Senate report cited in UK parliamentary evidence. A House of Commons Digital, Culture, Media and Sport Committee analysis found RT and Sputnik content had greater reach on Twitter for anti-EU material than the official Vote Leave and Leave.EU campaigns combined — meaning state-funded Russian media outperformed the UK’s own registered Leave campaign organizations on the platform where the referendum’s online battle was fought.
The UK Government’s Response: The 2020 “Russia Report”
The UK Parliament’s Intelligence and Security Committee investigated Russian interference in British politics, completing its findings in March 2019, though publication was delayed until July 2020 amid political controversy over the timing. The 47-page report concluded that Russian interference in UK politics was “the new normal” and that intelligence agencies had “actively avoided” formally investigating whether the Brexit referendum itself had been targeted — meaning the committee could not confirm or rule out impact, only that no one in government had seriously looked. The report did find evidence of coordinated Russian-linked activity following the 2014 Scottish independence referendum, establishing that the UK had already been a target of this playbook before the EU vote.
The Balkans Angle: Same Architecture, Different Theater
The organizational model used against Brexit — state broadcasters (RT, Sputnik) paired with covert social account networks (the IRA) — is the same architecture Russia later scaled into the Western Balkans. Sputnik Srbija, launched in 2015, became what regional researchers describe as often “the fastest, or the only,” wire source for Russia-friendly framing in Serbian-language media, allowing Kremlin narratives to be recycled into mainstream Balkan outlets much as RT and Sputnik’s English and European-language content was recycled into British social media conversation during the referendum. Analysts at the German Marshall Fund have documented the same narrative-proxy structure operating across Serbia, Bosnia and North Macedonia, where it was later deployed against NATO accession votes and the 2018 Macedonia name referendum.
Outcome: Unproven Impact, Proven Intent
No UK government body has concluded that Russian activity changed the referendum’s outcome, and the Intelligence and Security Committee was explicit that the absence of an investigation meant the question remained genuinely open rather than resolved in either direction. What is documented without dispute is intent and scale: a confirmed network of thousands of IRA-linked accounts, a broader universe of well over a hundred thousand Russia-linked accounts discussing the vote, and state broadcasters that outperformed the official campaigns they were ostensibly just covering. The 2020 report’s central finding — that nobody in UK government seriously checked — remains the case’s most consequential legacy for how European democracies now approach referendum security, and a cautionary example for every subsequent EU member state weighing whether to formally investigate its own vulnerable elections.
Leave a Reply