Russian interference in Brexit through targeted social media propaganda can be revealed for the first time. A cache of posts from 2016, seen by WIRED, shows how a coordinated network of Russian-based Twitter accounts spread racial hatred in an attempt to disrupt politics in the UK and Europe.
A network of accounts posted pro and anti-Brexit, anti-immigration and racist tweets around the EU referendum vote while also targeting posts in response to terrorist attacks across the continent. The accounts amplified their own messages to reach a greater audience and their impact raises questions about the full extent of Russia's propaganda campaign.
In a small snapshot of what is likely to be a much bigger issue, 139 tweets from 29 accounts show Russian trolls using hashtags related to the Brexit vote, pictures of London Mayor Sadiq Kahn, anti-Muslim language around European terror attacks and racial slurs against refugees.
The accounts, which were all confirmed as being Russian-backed by Twitter when it provided data to the US Congress, were followed by 268,643 people and some of their posts were retweeted hundreds of times. They were primarily created to disrupt the US presidential election but dabbled in wider issues around Brexit and European politics. All of the accounts have been suspended by Twitter.
Read more: Russian trolls live-tweeted Manchester and London attacks
The revelations come as a UK parliamentary inquiry begins its own investigations into Russian interference in the Brexit vote. Twitter and Facebook have both been asked to submit evidence of accounts directly related to the EU referendum and 2017's general election. Damian Collins, the leader of the parliamentary inquiry into fake news, says this first batch of tweets, though primarily focussed on US politics, "confirms what we've always believed" and that the accounts were trying to "influence political debate in the UK and also to insight hatred and turn communities against each other".
"I think it shows that Russian-controlled accounts have been politically active in the UK as well as America," Collins says. "This could just be the tip of the iceberg because we've only really just started looking and doing a proper detailed study of what accounts linked to Russian organisations have been doing politically."
The cache of tweets was provided by US security startup New Knowledge and were collected as part of a larger cache of information looking at extremism online.
When a Muslim woman was photographed crossing Westminster Bridge in the wake of a terror attack, an image went viral. Instead of showing her horror at the incident, the picture of her looking at her phone was taken out of context.
"Muslim woman pays no mind to the terror attack, casually walks by a dying man while checking phone #PrayForLondon #Westminster #BanIslam," tweeted the account @SouthLoneStar, the bio of which proclaimed the user was a "Proud TEXAN and AMERICAN patriot". The tweet was widely shared in news reports at the time.
@SouthLoneStar was identified as a Russian account by Twitter in response to its US inquiries about the country's influence in the 2016 presidential election. In June 2016, the account, which had 16,826 followers, also tweeted: "I hope UK after #BrexitVote will start to clean their land from muslim invasion!" and "UK voted to leave future European Caliphate! #BrexitVote". These posts were made after the referendum vote.
"The account occasionally wades into a European political discussion, which is not what I would expect a domestically-focused Conservative Texan to do under any circumstance," says Jonathon Morgan, CEO of New Knowledge. Morgan explains his team was researching alt-right behaviour on Twitter in the build-up to the US election, collecting 7,500 tweets from 40 accounts, all of which were run as Russian propaganda tools.
All the tweets, including those seen by WIRED, were gathered at the time they were posted using Twitter's API. The accounts appear in the Russian Twitter list published by the US Democrats. Some of the Tweets are archived by the Internet Archive, with unique user and tweet identification numbers provided by New Knowledge.
Twitter says it works proactively to stop "bots" from posting content, tries to check suspicious content, is working to improve how it detects single and cluster accounts created by suspicious sources and accounts that break its terms and conditions.
Surprisingly, all the posts around Brexit in this small snapshot were posted after the June vote. They included: "Brits said NO to prison of multiculturalism! Happy Independence Day!" (from @rightnpr) and "Let's hope that #Brexit will help Julian Assange. But I have a feeling that US won't let this happen" (@Jeblary2016, 8,054 followers).
"The accounts were focussed on posing as far-right activists in the US but they were using every opportunity they could to spread the pro-Kremlin message and its narrative," says Ben Nimmo, a senior fellow at the Atlantic Council, who has been researching Russian influence online and issues around the country's defence. "These Russian trolls are doing the same thing [as alt-right accounts], they are entirely in character but that's also the Kremlin narrative".
As well as focussing on European politics, one post from @SouthLoneStar reads, "France is turning into a shithole thanks to weak immigration laws and EU regulations" – there is a strong anti-Muslin and anti-Islam approach. After the July 2016 Munich terror attack, where nine people were killed the Russian account @Pamela_Moore13 tweeted, "Third attack in #Europe in 8 days. Multiple deaths in #Munich shooting. Europe is enjoying 'cultural enrichment'". @tpartynews (13,728 followers) tweeted "...Europe turns into Iraq! Very sad! #Munich #PrayForMunich".
Racial slurs were a popular theme: @PriceForPierce, which had 1,841 followers, tweeted, "ill welcome a European to the USA any day, refugees are not welcome in my eyes #IslamKills #StopIslam". And @ArchieOlivers (1,920 followers) wrote, "Have you seen the crimes many Syrian refugees R committing in Europe? #IslamKills #StopIslam".
After the terror attack on Brussels, which killed 31 people, @LeroyLovesUSA (1,201 followers) tweeted using the city's hashtag and #IslamKills: "Why can't EU just close the borders?" The message was sent on March 22 2016, the same day that so-called Islamic State claimed responsibility for the killings. A number of other Russian accounts posted similar messages at the time.
"There's this consistent and relentless focus on anti-immigrant sentiment, attacking Muslims in particular. [They are] making this sort of broad-brush, blanket link between Islam and terrorism," says Morgan, who collected the data and built a system to explore connections between accounts.
All the accounts seen by WIRED posted using either the Twitter web client of TweetDeck. The @Jenn_Abrams account had the most followers, 54,467, and the data pulled from the Twitter API shows some of the accounts were created as far back as 2013.
On both sides of the Atlantic, Twitter, Facebook and Google have come under-fire for allowing Russian-linked accounts to utilise their networks to run disinformation campaigns.
Senior representatives from the companies have been grilled by members of the US Congress on the extent of the problem. "I think that almost certainly when Twitter is pushed to reveal similar types of activity going on around Brexit, it will find a number of accounts that were behaving in a similar way," Morgan says, adding that he expects Russian propagandists specifically focussed on the UK to adopt right-wing British views, not right-wing American views.
At present, the UK inquiry into fake news is waiting to receive evidence back from Facebook and Twitter on UK-focussed Russian accounts. Collins, who is leading the inquiry, says both firms will take part in oral hearings by early-2018 at the latest. "I think we have a right to know if organisations in a foreign country, particularly in the case of Russia, are politically active," he says. "What is frightening, if you look at some of the studies that have come out of America, is just how many people a well run campaign can reach at a relatively low cost and how you can target and bombard people with highly partisan messages and fake news."
Separately, a study by academics at City University, reported on by BuzzFeed, found 13,000 Twitter bots were sending out pro-Brexit messages in the run-up to the vote. The bots were more likely to tweet pro-Leave rather than pro-Remain content. The Electoral Commission is also investigating whether Arron Banks, the prominent Leave campaigner, broke financial rules in the run-up to the vote as questions mount about where the money came from.
In the US, Russian adverts on Facebook reached 126 million Americans, almost half the country's population. On Twitter there were 2,752 accounts linked to Russia's Internet Research Agency and Google-owned YouTube found more than 1,000 videos from 18 accounts.
For Nimmo, this first glimpse at Russian interference in the Brexit vote is just the tip of the iceberg. "What are the people who were running those accounts doing now? "It's unlikely they've put their feet on the desk and said, 'We've lost that one, now let's go home.'"
This article was originally published by WIRED UK
