By Alexander J Martin, Technology Reporter
There is a lack of evidence supporting claims that the Kremlin used Facebook and Twitter to interfere in political debates in the West.
False personas and bots are openly visible on both social media platforms, primarily seeking to drive traffic to pornographic websites.
Facebook and Twitter’s failure to tackle the mass exploitation of their platforms for spam campaigns is a technical failure. There is certainly no reason to suspect that astroturfers (people presenting synthetic campaigns as if they were grassroots-led) would not be exploiting it, too.
But attempts to tie those astoturfing campaigns to an organised conspiracy by the Russian government have not been accompanied by key evidence.
It has now become a common allegation that both Facebook and Twitter were targeted by a private Russian company called the Internet Research Agency to influence the US presidential election, and potentially the Brexit referendum, supposedly on behalf of the Kremlin.
Despite initially resisting accusations of facilitating foreign disinformation campaigns, the social media companies have themselves begun to make these allegations.
Twitter submitted a list of 2,753 suspended accounts which it linked to the Internet Research Agency to a committee of US lawmakers. The committee is seeking evidence to support an unclassified intelligence community assessment that the Kremlin had orchestrated a “malign disinformation campaign targeting the 2016 presidential election”.
How the company established this list and those links is not known. Twitter has said they were identified via an anonymous non-governmental security-focused third party, but has refused to explain the methodology to Sky News, simply stating that the company had confidence in it.
Despite Twitter’s confidence, an explanation of that methodology is warranted. As reported by online technology magazine Motherboard, one of the suspended accounts – which has now been mysteriously reactivated – belongs not to an Internet Research Agency operative but to an American citizen called Robert Delaware, a man who says he has never even visited Russia.
Twitter declined to explain to Sky News why Mr Delaware’s account was presented to the House Intelligence Committee as being part of a Russian propaganda initiative, nor why it was subsequently reinstated.
The other 2,752 accounts remain suspended and no comprehensive archive of all of their posts and identities exists outside of Twitter’s data centres. There is no access to evidence to either support or challenge the merits of allegations that they were operatives in interference campaigns – and the public’s ability to make such assessments should be key.
Twitter’s own assessments of these accounts as being linked to the alleged troll factory is the primary source for every study providing details on organised attempts to influence the presidential election and Brexit. The studies do prove that such astroturfing campaigns took place, but they do not add any substance to allegations of Kremlin culpability – and there are other malicious actors who could have organised such campaigns.
If ridiculous, obviously fake information spreads – whether that’s the idea that MI5 would erase pencilled-in ballot papers or that the Democrats were running a paedophile ring from a pizza restaurant – that’s because we as a population are idiots. Even if a foreign power gives it a nudge, that level of stupidity is on us.
Astoturfing involving racist agitation, hate speech, political harassment and other forms of malicious communications have been mounted by online communities on 4chan and Reddit, as investigated by Dr Gianluca Stringhini of University College London.
Research by independent academics in Oxford and Edinburgh has identified organised astroturfing campaigns on Twitter, but the connection between suspect accounts and the Kremlin is held solely by the company.
Dr Stringhini told Sky News: “The research community doesn’t know where the accounts were connecting from, and this is something only Twitter can answer. From our point of view, these allegations of the accounts being connected to the Internet Research Agency are likely, but it is very difficult to get ground truth as researchers.”
The process of establishing ground truth using scientific evidence (evidence that can be reproduced and authenticated) is doubly valuable when it comes to fake personas and fake news, and not only for researchers. It is the standard of proof which would degrade the ability for politically charged lies to propagate. It is the panacea for our own idiocy.
Writing on social media companies and Russian influence campaigns, Sky’s Tom Cheshire said: “If ridiculous, obviously fake information spreads – whether that’s the idea that MI5 would erase pencilled-in ballot papers or that the Democrats were running a paedophile ring from a pizza restaurant – that’s because we as a population are idiots.
“Even if a foreign power gives it a nudge, that level of stupidity is on us.”
Just like Twitter, Facebook made similar claims to the committee, stating 80,000 posts made by Russian-based entities were seen by up to 126 million of its users ahead of the presidential election.
The copies of Facebook advertisements that have been made public are obviously constructed to be politically divisive, with some malicious actors even creating rallies for political opponents to be held in the same locations at the same time.
“We have seen fringe communities such as 4chan’s Politically Incorrect board or The Donald subreddit mount large-scale disinformation campaigns, creating conspiracy theories and pushing them into the mainstream by having social media talk about them, and reputable news outlets pick up on them,” Dr Stringhini said.
“At the moment we have anecdotes and single-case case studies showing particular disinformation activities, but we have no way of confirming who is behind this and whether it is a large-scale state sponsored campaign.”
This week, Facebook announced a bid to increase the transparency of those advertising with it by “creating a portal to enable people on Facebook to learn which of the Internet Research Agency Facebook Pages or Instagram accounts they may have liked or followed between January 2015 and August 2017”.
But the evidence of those ties to the Internet Research Agency are secret, and Facebook has refused to respond to enquiries about its methodology for identifying the culpable advertisements and pages.
Suspicions of fake personas and troll accounts have been mistargeted by more than just Twitter.
An article published through crowdfunded journalism site Byline Media articulated the author’s mistaken suspicion that the account of a Glaswegian security guard was actually operated by a Russian troll. The man behind the account ultimately turned to The Scotsman newspaper to reveal his identity and refute those claims.
At a meeting last week of senior ministers from EU nations, Spain’s foreign minister Alfonso Dastis expressed his government’s concern that Russia was behind social media posts agitating for Catalonian independence.
Spanish government ministers told national papers that claims of voters being pulled out of polling stations by their hair and stamped on by police were “fake news” spread by Russian trolls. These attacks were quite real and witnessed by millions through evidence posted on social media, and in person by Sky’s Europe Correspondent Mark Stone.
Discussing the value of evidence at an event titled The Automation of Consensus, held as part of Jesus College Cambridge’s new Intellectual Forum, Professor John Naughton cited the Gerasimov doctrine – written by the current chief of staff in Russia – which has been utilised during Russia’s annexation of the Crimea. It is a theory of military strategy which is reportedly prevalent in the Kremlin that emphasises the value of information warfare alongside kinetic operations.
The Gerasimov doctrine lends itself to deniability, but this does not mean that just because somebody can’t be linked to the Kremlin they secretly are. In the world of cybersecurity the term “patriotic hacker” has become used to describe hackers who are independently supporting the interests of the Russian state.
Corresponding with Sky News, Professor Mark Galeotti, an expert on Russian security affairs and researcher at the Institute of International Relations in Prague, said: “I am sure that the majority of pro-Russian trolls are not employees of the government or some private contractor agency, but people expressing their own views, so ‘patriotic trolls’ absolutely works as a concept.”
Dr Stringhini added: “Determining the provenance of information is very important here. Social network users should be provided with information on where a certain news story originated, who pushed it, and how it reached the user. In this context, whether a misinformation campaign was created by organised trolls or by a state-sponsored actor makes a difference, both from the user awareness perspective and from the national security one.”
Most of the questions we can direct at social media companies and governments can and should be responded to with definite answers and reproducible evidence, because most of the people working for social media companies and governments are intelligent enough to recognise the difference between their authority and the truth.
Regarding these misinformation campaigns, an unclassified intelligence community assessment in the US has pointed the finger squarely at Russia and Vladimir Putin. That attribution requires evidence. Not because its hypothesis is unlikely, but because it is likely. Because the integrity we show when identifying a disinformation campaign should not reflect the methodology of that campaign and tell us what we already know to be true, but rather set a standard of evidence we can deploy when we encounter these campaigns again in the future.