Researchers 'don't know' who was behind Russian Twitter bots that churned out pro-Corbyn posts, or how much difference it made

John McDonnell dismisses report as 'classic Sunday Times smear campaign'

Labour leader Jeremy Corbyn on the general election campaign trail in Warrington (PA)

Researchers say they do not know who was behind Russia-linked Twitter “bots” that churned out messages supporting Jeremy Corbyn and attacking the Conservatives in the run-up to the general-election.

Matt Hancock, the digital and culture secretary, expressed concern over the suggestion that state actors were behind automated accounts that shared praise of Labour policies, details of rallies and urged Mr Corbyn’s supporters to turn out on polling day.

Responding to analysis by Swansea University and The Sunday Times, the Tory minister said it was “absolutely unacceptable for any nation to attempt to interfere in the democratic elections of another country”.

But experts cautioned that it was impossible to prove who was behind the networks, their motives or what impact they had in the “murky” world of information warfare.

John McDonnell, the shadow chancellor, dismissed the assertion that Russians “tried to swing the election” in Mr Corbyn's favour as a “classic Sunday Times smear campaign”.

“This is ludicrous,” he told Sky News. “If I remember rightly, the Russian Embassy was putting out supportive noises towards the Tory party.

“If there’s an issue here about anything with Russian influence in our society, it’s about Russian oligarchs funding the Tory party – let’s have an inquiry into that.”

Oleksandr Talavera, a professor of finance at Swansea University who researches social media in elections, monitored thousands of tweets using hashtags related to the 2017 general election, political parties and leaders.

In the two months either side of the 8 June vote, he found more than one million posts in support of Labour but fewer than 388,000 backing the Conservatives and other parties far behind.

When looking only at accounts that tweeted in English but appeared to originate in Russia, he found 21,448 tweets on British politics were created in 30 days leading up to the election.

“Most Labour-hashtagged tweets were positive but most Tory ones were negative, and that’s interesting coming from Russian accounts – it indicates some sort of direction,” Professor Talavera told The Independent.

“It is difficult to know how many people saw the tweets, but we are only looking at 1 per cent of all tweets so theoretically the numbers can be multiplied up. As for who is behind this, we don’t know.”

He pointed at a study suggesting that sentiments expressed on Twitter could affect the stock market as an indicator of the potential impact but that there was currently no measure of how tweets affected voting.

Many of the Twitter bots studied by The Sunday Times shared 15-character alphanumeric usernames and English-sounding women’s names, despite listing their first language as Russian.

Much of their activity was conducted in unison, retweeting official Labour accounts, Momentum and unions.

Around 80 per cent of the automated accounts had been created after Theresa May called the snap election and became more active at key points including the launch of the Conservative Party manifesto, Manchester Arena bombing and television debates.

The vast majority of accounts have now been suspended or removed by Twitter and it is impossible to view their full output, but it seems some were not exclusively political.

Twitter works to detect and remove spam accounts (AFP/Getty)

One bot in the survey, under the name “Denis”, also sent out pornographic images and positive messages about Donald Trump.

Ben Nimmo, a fellow at the Atlantic Council's Digital Forensic Research Lab, told The Independent that the dates the accounts were created pointed to their purpose but they could be used for several aims.

“For me the key question is: are these commercial bots that someone hired to amplify political messaging in the UK, or are they dedicated bots that someone set up to generate the political effect themselves?” he asked.

“What we’ve seen before is Russian-language botnets [networks] that start posting on different themes but are primarily there to make money for the bot herder [manager].”

Bots can be used for a variety of purposes, including monitoring, marketing or allowing people to buy fake social media followers.

Others are governed simply by traffic and interaction, causing them to share tweets on topics that are dominating social media at any given time, as Labour was during the general election.

Owners of the networks can be paid to change what bots are programmed to do, seeing some switch between advertising cars and jewellery to political parties, for example.

Even the Russian government’s own Internet Research Agency, known as the “troll factory” for its influence operations, periodically hires external botnets to amplify chosen messages.

“It’s a very murky world and the nature of data from Twitter means you can say ‘this is probably a Russian botnet’, but to go from that to saying who is responsible is very difficult,” Mr Nimmo explained.

“If you look into the data enough you can get a probability, [but] it’s very hard to get a certainty.”

Professor Talavera said that although Twitter does a “good job” at removing malicious bots, it could help users distinguish tweets by flagging whether they were automated and increasing transparency around advertising.

A Twitter spokesman said the firm was working to improve its own systems and to educate the public.

“Our work to fight both malicious bots and misinformation goes beyond any one specific election, event, or time period,” he added.

“We’ve spent years working to identify and remove spammy or malicious accounts and applications on Twitter. We continue to improve our internal systems to detect and prevent new forms of spam and malicious automation.”

The research comes following persistent allegations of Russian influence in foreign elections, which the Kremlin denies, and wider cyber warfare.

An unprecedented joint technical alert issued by British and American security services earlier this month said Russian hackers were targeting millions of devices around the world to spy, steal information and build networks for potentially devastating future cyberattacks.

The government has announced £150m funding to bolster the NHS’s cyber defences after the devastating WannaCry virus, and intelligence agencies say other parts of Britain’s critical national infrastructure, including energy networks, have been targeted in the past.