237.The speed of technological development has coincided with a crisis of confidence in institutions and the media in the West. This has enabled foreign countries intent on destabilising democratic institutions to take advantage of this crisis. There has been clear and proven Russian influence in foreign elections, and we highlighted evidence in our Interim Report of such attempts in the EU Referendum.264
238.It is interesting to note that, as of 30 November 2018, the online Government response to our Report received a total of 1,290 unique page views and the PDF has been visited 396 unique times from the website.265 In the month following its publication, over 63% of views of the report online were from foreign IP addresses (whereas, on average, 80% of viewers of Reports are UK-based), and of these, over half were from Russia. Furthermore, two-thirds of viewers were new visitors, meaning they had not visited the parliament.uk website before (in comparison with the majority of Reports, where only around 30% are new visitors). The following table shows the unique page views by city, illustrating this high proportion from Russia:
Source: Web and publications Unit, House of Commons
The following map shows the concentration of those readers of the Government Response to the Interim Report, by country:
Source: Web and Publications Unit, House of Commons
This itself demonstrates the very clear interest from Russia in what we have had to say about their activities in overseas political campaigns.
239.In this Chapter, we will update the information we set down in our Interim Report, including Facebook’s knowledge about Russian interference in its data. We shall also build on our previous recommendations.
240.As we said in our Interim Report, Prime Minister Theresa May accused Russia of meddling in elections and planting disinformation, in an attempt to ‘weaponise information’ and sow discord in the West.266 In its response to the Report, the Government stated that, following the nerve agent attack in Salisbury in March 2018, the Government had “judged the Russian state promulgated at least 38 false disinformation narratives around this criminal act”.267 However, the Government made it clear that “it has not seen evidence of successful use of disinformation by foreign actors, including Russia, to influence UK democratic processes”.268
241.When the Secretary of State was questioned in oral evidence over what constitutes “successful”, Rt Hon Jeremy Wright MP, responded: “We have seen nothing that persuades us that Russian interference has had a material impact on the way in which people choose to vote in elections. It is not that they have not tried, but we have not seen evidence of that material impact”.269 It is surely a sufficient matter of concern that the Government has acknowledged that interference has occurred, irrespective of the lack of evidence of impact. The Government should be conducting analysis to understand the extent of Russian targeting of voters during elections.
242.The Government also cannot state definitively that there was “no evidence of successful interference” in our democratic processes, as the term “successful” is impossible to define in retrospect. There is, however, strong evidence that points to hostile state actors influencing democratic processes. Cardiff University and the Digital Forensics Lab of the Atlantic Council have both detailed ways in which the Kremlin attempted to influence attitudes in UK politics.270
243.Kremlin-aligned media published significant numbers of unique articles about the EU referendum. 89 Up researchers analysed the most shared of the articles, and identified 261 with a clear anti-EU bias to the reporting. The two main outlets were RT and Sputnik, with video produced by Ruptly.271 The articles that went most viral had the heaviest anti-EU bias.272 The social reach of these anti-EU articles published by the Kremlin-owned channels was 134 million potential impressions, in comparison with a total reach of just 33 million and 11 million potential impressions for all content shared from the Vote Leave website and Leave.EU website respectively.273 The value for a comparable paid social media campaign would be between £1.4 and 4.14 million.
244.On 17 January 2019, Facebook removed 289 Pages and 75 accounts from its site, accounts that had about 790,000 followers and had spent around $135,000 on ads between October 2013 and January 2019. The sites had been run by employees at the Russian state-owned news agency Sputnik, who represented themselves as independent news or general interest Pages. Around 190 events were hosted by these Pages (the first was scheduled for August 2015 and the most recent was scheduled for January 2019).274
245.Nathaniel Gleicher, Facebook’s head of cybersecurity policy, wrote: “Despite their misrepresentations of their identities, we found that these Pages and accounts were linked to employees of Sputnik, a news agency based in Moscow, and that some of the Pages frequently posted about topics like anti-NATO sentiment, protest movements, and anti-corruption.”275 Facebook also removed 107 Pages, groups and accounts that were designed to look as if they were run from Ukraine, but were part of a network that originated in Russia.
246.Ben Nimmo, from the Digital Forensics Lab of the Atlantic Council, has detailed attempts to influence attitudes to the Scottish Referendum, for instance, which included a Russian election observer calling the referendum not in line with international standards, and Twitter accounts calling into question its legitimacy. The behaviour of these accounts, Mr Nimmo argues, is pro-Kremlin, and consistent with the behaviour of accounts known to be run by the so-called “troll factory” in St. Petersburg, Russia, during the United States 2016 presidential election and beyond. However, it is not possible to determine from open sources whether some or all of the accounts are independent actors, or linked to Russian information operations.276
247.As the Secretary of State said, Russia also used malign digital influence campaigns to undermine the Government’s communication of evidence in the aftermath of the poisoning of the Skripals.277 Research by the Centre for Research and Evidence on Security Threats at Cardiff University showed how ‘sock puppet’ Twitter accounts,278 controlled by the St Petersburg-based ‘Internet Research Agency’, tried to fuel social divisions, including religious tensions, in the aftermath of the Westminster, Manchester, London Bridge and Finsbury Park terror attacks.279 Furthermore, the methods through which malign influence can be deployed are also constantly being expanded. While Twitter has been responsive in shutting down abusive and fake accounts, Facebook remains reluctant to do so. Research by the Institute for Strategic Dialogue and the LSE Arena Program into the German 2017 elections discovered Facebook Groups created by unverifiable administrators, directing Russian state-backed media during the election period, with regularity, across social media.280
248.The Government has been very ready to accept the evidence of Russian activity in the Skripal case, an acceptance justified by the evidence. However, it is reluctant to accept evidence of interference in the 2016 Referendum in the UK. If the Government wishes the public to treat its statements on these important matters of national security and democracy seriously, it must report the position impartially, uninfluenced by the political implications of any such report.
249.In common with other countries, the UK is clearly vulnerable to covert digital influence campaigns and the Government should be conducting analysis to understand the extent of the targeting of voters, by foreign players, during past elections. We ask the Government whether current legislation to protect the electoral process from malign influence is sufficient. Legislation should be in line with the latest technological developments, and should be explicit on the illegal influencing of the democratic process by foreign players. We urge the Government to look into this issue and to respond in its White Paper.
250.The Committee has repeatedly asked Facebook, in written correspondence and in oral evidence, about Russian activity on Facebook and, in particular, about knowledge of the Russian adverts that ran during the presidential election in America in 2016. According to a New York Times article published in November 2018,281 Facebook had discovered suspicious Russia-linked activity on its site in early 2016, in an attempt to disrupt the presidential election. In September 2017, Alex Stamos, the then Chief Security Officer, told the members of Facebook’s Executive Board that that Russian activity was still not under control. The article claimed that Facebook executives, including Mark Zuckerberg and Sheryl Sandberg, attempted to deflect attention from their own company to other tech companies, by hiring a political consultancy, Definers Public Affairs, allegedly to spread anti-semitic information about George Soros and his campaigning activities, after Mr Soros called Facebook “a menace to society” in early 2018.282
251.When Simon Milner, Policy Director UK, Middle East and Africa, at Facebook, gave evidence to us in February 2018, he was asked specifically about whether Facebook had experienced people from one country seeking to place political adverts in another country. He replied:
We have not seen in the last general election, during the Brexit vote or during the 2015 general election, investigative journalism, for instance, that has led to the suggestion that lots of campaigns are going on, funded by outsiders. […] There is no suggestion that this is going on.283
252.Given the information contained in the New York Times article and the information we have received from Six4Three, we believe that Facebook knew that there was evidence of overseas interference and that Mr Milner misled us when he gave evidence in February 2018. Facebook’s Chief Technology Officer, Mike Schroepfer, also told the Committee, with regards to the company’s knowledge of Russian interference in the 2016 presidential election, by targeting user accounts on the site: “We were slow to understand the impact of this at the time”.284 Again, this would appear to be a misleading answer based on what senior executives at Facebook knew in 2016. We now know that this statement was simply not true. We are left with the impression that either Simon Milner and Mike Schroepfer deliberately misled the Committee or they were deliberately not briefed by senior executives at Facebook about the extent of Russian interference in foreign elections.
253.The Six4Three documents revealed that “an engineer at Facebook notified the company in October 2014 that entities with Russian IP addresses had been using a Pinterest API key to pull over 3 billion data points a day through the ordered friends API. This activity was not reported to any external body at the time”.285
254.When questioned about these emails, Richard Allan refused to answer, stating that that information was based on emails that were “unverified, partial accounts from a source who has a particular angle”.286 However, on the same evening of 27 November 2018, Facebook itself chose to send the very same emails to a CNN Reporter, despite Richard Allan’s description of them..287 Facebook wanted to show that the investigation had proved that there had been no Russian interference. However, the email exchange shows that the engineer’s reassurance of there being no Russian interference was given within an hour, and it is questionable whether Facebook engineers would have been able to satisfy themselves within that short time that Russian interference had not occurred.
255.The ICO has been investigating how far data was shared between GSR—the company set up by Dr. Kogan, in advance of his work involving the ‘thisisyourdigitallife’ app—Cambridge University, and Russian APIs.288 When asked whether the ICO was still investigating whether Dr Kogan’s data had been accessed by people in Russia, the Information Commissioner, Elizabeth Denham replied:
It is an active line of investigation. What we said in July was that there were some IP addresses that were found in that data and that server associated with Aleksandr Kogan that resolved to Russia and associated states. That is information that we have passed on to the authorities. It is not in our remit to investigate any further than that, but we have passed that on to the relevant authorities.289
She later told us that the ICO had referred the issue to the National Crime Agency.290
256.Further clarification from the Deputy Information Commissioner, James Dipple-Johnston, highlighted the fact that IP addresses originating from Russia were connected to an earlier app at the Cambridge University Psychometrics Centre. The IP addresses were also linked to alleged cyber attacks in the past and to a “Tor entry point”—a device for people to hide their identity online.291
257.Russian meddling in elections overseas has, clearly, not been limited to just Facebook. In October 2018, Twitter released an archive of tweets that had been shared by accounts from the Internet Research Agency, with the goal of “encouraging open research and investigation of these behaviors from researchers and academics around the world”.292 The datasets comprised of 3,841 accounts that were affiliated with the Internet Research Agency and originated from Russia, and 770 other accounts, “potentially originating in Iran”. The accounts included more than 10 million tweets and more than 2 million images.293 The Twitter accounts were used to influence the 2016 US presidential election, as well as elections and referenda in several other countries, including the UK. The accounts were also used to influence public sentiment around several issues of national importance in other countries, including Ukraine.
258.The Oxford Internet Institute and the Senate Select Committee on Intelligence worked together to inquire into the activities of the Internet Research Agency (IRA), by studying data that had been provided by the tech companies in the summer of 2017.294 The investigations revealed that: the Russian campaign to polarise the US electorate and destabilise trust in the media started in 2013, which is earlier than previously thought; and the IRA subsequently accelerated content production across a full set of social media companies, with parallel trends across Twitter, Facebook, Instagram and YouTube.
259.We note as well the comments made by Vladislav Surkov, a senior advisor to President Putin, in an article published in the Russian daily Nezavisimaya Gazeta, on 11 February 2019. He said that “Foreign politicians blame Russia for meddling in elections and referenda all over the planet. In fact, it’s even more serious than that: Russia is meddling in their brains and they don’t know what to do with their changed consciousness.”.295
260.Our Interim Report highlighted the fact that Arron Banks is, to date, the person who has, allegedly, given the largest donation to a political campaign in British history, reported to be £8.4 million, but that questions still remain over both the sources of that donation and the extent of Mr Banks’ wealth.296
261.The Report recorded the fact that Arron Banks discussed business ventures within Russia and beyond, in a series of meetings with Russian Embassy staff:
Arron Banks and Andy Wigmore have misled the Committee on the number of meetings that took place with the Russian Embassy and walked out of the Committee’s evidence session to avoid scrutiny of the content of the discussions with the Russian Embassy. […] It is unclear whether Mr. Banks profited from business deals arising from meetings arranged by Russian officials.297
262.Our Interim Report recommended that the Electoral Commission pursue investigations into donations that Arron Banks made to the Leave campaign, to verify that that money was not sourced from abroad, and that “should there be any doubt, the matter should be referred to the National Crime Agency”.298 On 1 November 2018, the Electoral Commission referred the following organisations and individuals to the National Crime Agency: Better for the Country (the company that ran the Leave.EU referendum campaign); Arron Banks; Leave.EU; Elizabeth Bilney; and other associated companies and individuals. The Electoral Commission’s investigation focused on £2m reported to have been loaned to Better for the Country by Arron Banks and his group of insurance companies and a further £6m reported to have been given to the organisation, on behalf of Leave.EU, by Arron Banks alone.299 The NCA has now launched a criminal investigation.
263.We asked the National Crime Agency for an update on their investigations and they replied:
The NCA has initiated an investigation concerning the entities Better for the Country (BFTC) and Leave.EU; as well as Arron Banks, Elizabeth Bilney and other individuals. This follows our acceptance of a referral of material from the Electoral Commission. This is now a live investigation, and we are unable to discuss any operational detail.300
264.In the spring of 2018, we heard that Steve Bannon had introduced Arron Banks to Cambridge Analytica.301 In November 2018, we received evidence to show that there was a relationship between Leave.EU and Steve Bannon in 2015, highlighted in an email from Arron Banks to Andy Wigmore, copying in Steve Bannon and Elizabeth Bilney, showing that Leave.EU wanted Cambridge Analytica to set up a funding strategy in the US:
We would like Cambridge Analytica to come up with a strategy for fundraising in the States […] and how we could connect to people with family ties in the UK and raise money and create SM [social media] activity.302
Arron Banks and Leave.EU had not only Russia, but the US, in their sights.
265.The Electoral Commission’s paper, “Digital Campaigning”, published in June 2018, highlights the fact that the current rules on spending were established in a pre-digital time:
The UK’s rules set minimum amounts for campaign spending before people or organisations have to register as a non-party campaigner. This means that a foreign individual or organisation that spends under these amounts would not have broken any specific electoral laws in the UK. […] Had not seen potential for foreign sources to direct purchase campaign advertising in the UK”.303
266.We are pleased that our recommendation set out in the Interim Report in July 2018, concerning Arron Banks and his donation, has been acted on by both the Electoral Commission—which has concerns that Banks is not the ‘true source’ of the donation—and by the National Crime Agency, which is currently investigating the source of the donation.
267.There is a general principle that, subject to certain spending limits, funding from abroad is not allowed in UK elections. However, as the Electoral Commission has made clear, the current rules do not explicitly ban overseas spending. We recommend that, at the earliest opportunity, the Government reviews the current rules on overseas involvement in our UK elections to ensure that foreign interference in UK elections, in the form of donations, cannot happen. We also need to be clear that Facebook, and all platforms, have a responsibility to comply with the law and not to facilitate illegal activity.
268.We have recently been told by Clint Watts, an expert whom we first met in New York in February 2018, that Twitter accounts monitored during 2015 were discussing both Brexit and the US Presidential campaign influence.304 Furthermore, in the same month that Twitter released its archive of tweets shared by accounts from the Internet Research Agency, the US Department of Justice filed criminal charges against a Russian national, Elena Alekseevna Khusyaynova, for alleged crimes relating to interference between the period of the 2016 US presidential election and the 2018 mid-term elections.305
269.The FBI also filed a Criminal Complaint on 28 September 2018. It described the work of ‘Project Lakhta’, in which individuals have allegedly “engaged in political and electoral interference operations targeting populations within the Russian Federation and in various other countries, including, but not limited to, the United States, members of the European Union, and Ukraine”.306 Since at least May 2014, Project Lakhta’s stated goal in the United States was to spread distrust towards candidates for political office and the political system in general”.307 The complaint also listed 14 companies—believed to be shell companies—involved in the conspiracy.308
270.As recently as 31 January 2018, Facebook announced the suspension of a network of accounts—783 pages, groups, and accounts—that it said were engaged in coordinated inauthentic behaviour on Facebook and Instagram that was “directed from Iran.” Almost simultaneously, Twitter announced that it had suspended networks of accounts that it termed “foreign information operations”, potentially connected to Iran, Venezuela and Russia.309
271.Information operations are part of a complex, interrelated group of actions that promote confusion and unrest through information systems, such as social media companies. These firms, in particular Facebook, need to take action against untransparent administrators of groups, which are being used for political campaigns. They also need to impose much more stringent punishment on users who abuse the system. Merely having a fake disinformation account shut down, but being able to open another one the next moment, is hardly a deterrent.
272.The Government should put pressure on social media companies to publicise any instances of disinformation. The Government needs to ensure that social media companies share information they have about foreign interference on their sites—including who has paid for political adverts, who has seen the adverts, and who has clicked on the adverts—with the threat of financial liability if such information is not forthcoming. Security certificates, authenticating social media accounts, would ensure that a real person was behind the views expressed on the account.
273.We repeat our call to the Government to make a statement about how many investigations are currently being carried out into Russian interference in UK politics. We further recommend that the Government launches an independent investigation into past elections—including the UK election of 2017, the UK Referendum of 2016, and the Scottish Referendum of 2014—to explore what actually happened with regard to foreign influence, disinformation, funding, voter manipulation, and the sharing of data, so that appropriate changes to the law can be made and lessons can be learnt for future elections and referenda.
264 Disinformation and ‘fake news’: Interim Report, DCMS Committee, Fifth Report of Session 2017–19, HC 363, 29 July 2018, Chapter 5.
265 These statistics have been supplied by the Web and Publications Unit, House of Commons. ‘Unique’ means that if the same person visited a HTML page/PDF multiple times in one session it would count as one view only). It is not possible to log any reads of the PDF which have not come from the Parliament.uk website (for example, when the link to the PDF is shared on Twitter) so this statistic is deceptively low.
266 Disinformation and ‘fake news’: Interim Report, DCMS Committee, Fifth Report of Session 2017–19, HC 363, 29 July 2018, para 41.
267 Disinformation and ‘fake news’: Government Response to the Committee’s Fifth Report of Session 2017–19, 23 October 2018, HC 1630 Government response to Interim Report, page 16.
268 Same as above.
269 Q211 Evidence session, 24 October 2018, The Work of the Department for Digital, Culture, Media and Sport.
270 Russian influence and interference measures following the 2017 UK terrorist attacks, Cardiff University Crime and Security Research Institute, funded by Centre for Research and Evidence on Security Threats (CREST), 18 December 2017; #Election Watch: Scottish vote, pro-Kremlin trolls: how pro-Russian accounts boosted claims of election fraud at Scotland’s independence referendum, DFRLab, 12 December 2017.
271 Ruptly GmbH is a video news agency that is owned by the RT televised news network.
272 89up releases report on Russian influence in the EU referendum, 89up, 2 October 2018, slide 10.
273 89up releases report on Russian influence in the EU referendum, 89up, 2 October 2018.
274 Removing coordinated inauthentic behavior from Russia, Facebook newsroom, 17 January 2019.
275 Same as above.
276 Russians ‘tried to discredit 2014 Scots independence vote’, Chris Marshall, 13 December 2017; #Election Watch: Scottish Vote, Pro-Kremlin Trolls, medium, 13 December 2017
277 Q215, Rt Hon Jeremy Wright QC MP evidence session, 24 October 2018, The Work of the Department for Digital, Culture, Media and Sport, HC 361.
278 A sockpuppet is an online identity used for purposes of deception.
279 Russian influence and interference measures following the 2017 UK terrorist attacks, Cardiff University Crime and Security Research Institute, funded by Centre for Research and Evidence on Security Threats (CREST), 18 December 2017.
280 “Make Germany great again”: Kremlin, alt-right and international influences in the 2017 German elections, Institute for Strategic Dialogue and the Institute of Global Affairs, December 2017.
281 Delay, Deny and Deflect: How Facebook’s Leaders Fought Through Crisis, Sheera Frenkel, Nicholas Confessore, Cecilia Kang, Matthew Rosenberg and Jack Nicas, The New York Times, 14 November 2018.
287 Donie O’Sullivan Twitter account (@donie) , incorporating the redacted Facebook emails, 27 November 2018.
288 Investigation into the use of data analytics in political campaigning: a report to Parliament, ICO, 6 November 2018.
292 Enabling further research on information operations on Twitter, Vijaya Gadde and Yoel Roth, Twitter website, 17 October 2018.
293 Same as above.
294 The IRA, Social Media, and Political Polarization in the United States, 2012–2018, Philip N.Howard, Bharath Ganesh, Dimitra Liosiou, University of Oxford, and John Kelly, Camille Francois, Graphika, 18 December 2018.
295 Official: Russia’s political system a good model for others, Vladimir Isachenkov, The Washington Post, 11 February 2019.
296 Disinformation and ‘fake news’: Interim Report, DCMS Committee, Fifth Report of Session 2017–19, HC 363, 28 July 2018, paras 187 to 188.
297 Disinformation and ‘fake news’: Interim Report, DCMS Committee, Fifth Report of Session 2017–19, HC 363, 28 July 2018, paras 185 and 186.
298 Same as above, para 191.
299 Report on investigation into payments made to Better for the Country and Leave.EU, Electoral Commission, 1 November 2018.
300 Email sent to the Committee, 16 November 2018.
303 Digital campaigning: increasing transparency for voters, The Electoral Commission, June 2018, para 86.
304 Private conversation with Clint Watts, Distinguished Research Fellow at the Foreign Policy Research Institute.
305 Criminal complaint, USA v Elena Alekseevna Khusyaynova, case No. 1:18-MJ-464, District Court Alexandria, Virginia, 28 September 2018.
306 Criminal complaint, USA v Elena Alekseevna Khusyaynova, case No. 1:18-MJ-464, District Court Alexandria, Virginia, 28 September 2018.
307 Same as above.
308 Same as above.
309 Facebook and Twitter remove thousands of fake accounts tied to Russia, Venezuela and Iran, Donie O’Sullivan, CNN Business, 31 January 2019.
Published: 18 February 2019
