EUROPEAN COMMISSION

Brussels, 14.6.2019

JOIN(2019) 12 final

JOINT COMMUNICATION TO THE EUROPEAN PARLIAMENT, THE EUROPEAN COUNCIL, THE COUNCIL, THE EUROPEAN ECONOMIC AND SOCIAL COMMITTEE AND THE COMMITTEE OF THE REGIONS

Report on the implementation of the Action Plan Against Disinformation

A.Introduction

Protecting our democratic processes and institutions from disinformation 1 is a major challenge for our societies. In order to tackle this challenge, the EU has put in place a robust framework for coordinated action which is fully in line with our European values and fundamental rights.

European democracy is only as strong as the active participation of its citizens. The record high turnout in the recent European Parliament elections shows how engaged Europeans want to be in shaping the future of their Union.

While it is too early to draw final conclusions about the level and the impact of disinformation on these elections, it is clear that the measures taken as part of the Joint Action Plan against Disinformation 2 and the dedicated Elections Package 3 contributed to deter attacks and expose disinformation. Encouraged by these actions, numerous journalists, fact checkers, platforms, national authorities, researchers and civil society contributed to raising awareness about how to counter the threat. Increased public awareness made it harder for malicious actors to manipulate the public debate.

However, there is no room for complacency and the fight against disinformation must continue. It is a long-term challenge that concerns all parts of our societies and requires continuous commitment and efforts. More needs to be done to protect our Union’s democratic processes and institutions from manipulation and disinformation.

Today’s report by the Commission and the High Representative provides a first assessment of the progress achieved so far and sets out the main lessons for the future. It explains in more detail how the Action Plan and the Elections Package helped to fight disinformation in the context of the European elections. It is also the contribution of the Commission and the High Representative to the European Council meeting on 20-21 June 2019.

B.Towards a coordinated approach to tackle disinformation

With the adoption of the Action Plan against Disinformation, the Commission and the High Representative put in place a strong framework to counter threats inside and outside of the EU. All relevant actors, including EU institutions, Member States, industry and civil society played their respective role along four strands of action:

1.The EU has strengthened its capabilities to identify and counter disinformation, via the Strategic Communication Task Forces and the EU Hybrid Fusion Cell in the European External Action Service. It has also improved the coordinated response by setting up a Rapid Alert System to facilitate the exchange of information between Member States and the EU institutions.

2.The EU worked with online platforms and industry through a voluntary Code of Practice on disinformation to increase transparency of political communications and prevent the manipulative use of their services. This allows users to know why they see specific political content and ads, and see where they come from and who is behind them.

3.The Commission and the High Representative, in cooperation with the European Parliament, helped increase awareness and resilience to disinformation within society, notably through more fact based messaging and renewed efforts to promote media literacy.

4.The Commission has supported Member States’ efforts to secure the integrity of elections and strengthen the resilience of the Union’s democratic systems, by facilitating collaboration, providing guidance and support, and through legislative measures.

1.Improving capabilities and strengthening coordinated responses

The Commission and the High Representative strengthened the Union’s capabilities to detect, analyse and expose disinformation, as well as to ensure a coordinated response, in particular through the Rapid Alert System. This was notably done by reinforcing related work strands in the Directorate General for Communication and by stepping up the financial and human resources of the Strategic Communication Task Forces of the European External Action Service. For instance, the increased budget helped the East Strategic Communications Task Force to expand the scope of their activities and further develop professional monitoring.

The Rapid Alert System facilitated daily exchanges and sharing of information on a number of cases and trends related to disinformation between EU authorities and Member States.

At this point in time, available evidence has not allowed to identify a distinct cross-border disinformation campaign from external sources specifically targeting the European elections. However, the evidence collected revealed a continued and sustained disinformation activity by Russian sources 4 aiming to suppress turnout and influence voter preferences. These covered a broad range of topics, ranging from challenging the Union’s democratic legitimacy to exploiting divisive public debates on issues such as of migration and sovereignty. This confirms that the disinformation campaigns deployed by state and non-state actors 5  pose a hybrid threat to the EU.

There was a consistent trend of malicious actors using disinformation to promote extreme views and polarise local debates, including through unfounded attacks on the EU. Domestic political actors often adopted the tactics and narratives used by Russian sources to attack the EU and its values 6 . Other external actors were also involved.

The tactics used by these actors are evolving as quickly as the measures adopted by states and online platforms. Instead of conducting large-scale operations on digital platforms, these actors, in particular linked to Russian sources, now appeared to be opting for smaller-scale, localised operations that are harder to detect and expose.

Given the increasingly sophisticated nature of disinformation activities, and the difficulties of independent researchers to access relevant data from the platforms, a conclusive assessment of the scope and impact of disinformation campaigns will take time and require a concerted effort by civil society, academia, public actors and online platforms.

The Rapid Alert System has demonstrated its value by fostering closer coordination among EU institutions and national authorities. The number of interactions between authorities has been growing steadily and the tool has become a reference point for the fight against disinformation. It has also strengthened cooperation with online platforms, although the platforms still need to become more responsive to external reports of inauthentic behaviour and malign content. The Rapid Alert System also facilitated cooperation with international partners, such as the G7 and the North Atlantic Treaty Organization, and this will be further strengthened in the future.

2.Implementation of the Code of Practice on Disinformation

Online platforms have become powerful gateways to information for many EU citizens. They therefore can play a key role in the spread of disinformation. This is why the Commission has urged platforms to step up their efforts in the fight against disinformation. As a result, the main platforms, including Facebook, Google, Twitter, along with software companies and bodies representing the advertising industry adopted in October 2018 a self-regulatory Code of Practice on disinformation 8 . In doing so, they voluntarily committed to improve the transparency, accountability and trustworthiness of their services.

In view of the 2019 European elections, the Commission and the European Regulators Group for Audiovisual Media Services (ERGA) 9 carried out a targeted monitoring of the actions taken by Facebook, Google and Twitter on the basis of monthly reports submitted by these platforms from January to May 2019 10 . This monitoring identified the following achievements:

·The online platforms improved their scrutiny of ad placements to limit malicious click-baiting practices and reduce advertising revenues for those posting disinformation. Their monthly reports provided detailed figures, by Member State, on the number of ads taken down and ad accounts closed due to deceptive or misleading behaviour 11 .

·All three online platforms stepped up their efforts to improve transparency for political ads, in particular by labelling them and making them publicly available via searchable ad libraries. 

·The platforms reported to have taken actions to support the integrity of their services. They reported to focus on manipulative behaviour to make content more visible through coordinated operations, as well as the abusive use of bots and fake accounts. 

In the days preceding the elections, online platforms identified and took down additional accounts spreading disinformation and hate speech following alerts from independent investigators and journalists 12 . More than 600 groups and Facebook pages operating across France, Germany, Italy, the United Kingdom, Poland and Spain were reported to have spread disinformation and hate speech or have used false profiles to artificially boost the content of parties or sites they supported. These pages generated 763 million user views. Reports from researchers, fact-checkers and civil society also identified additional instances of large-scale attempts to manipulate voting behaviour across at least nine Member States 13 .

Notwithstanding these efforts, more needs to be done by the platforms to effectively tackle disinformation. Although Facebook extended its transparency to issue-based ads 14 and Google and Twitter did not, questions remain about the effectiveness of the transparency measures taken by all signatories. Furthermore, the platforms did not make sufficient progress in increasing the transparency of websites hosting ads, partly due to the lack of engagement from the advertising industry.

All platforms should also ensure an active and working cooperation with fact checkers in all Member States and empower users to better detect disinformation. This can be done, for instance, by developing trustworthiness indicators for information sources in cooperation with media organisations.

Platforms should also give meaningful access to data to the research community, in line with personal data protection rules. The cooperation with researchers will enable a better detection and analysis of disinformation campaigns, a sound monitoring of the Code’s implementation and independent oversight of the functioning of algorithms. The Commission will continue to promote a wide take-up of the Code amongst all relevant stakeholders.

Before the end of the year, the Commission will assess the effectiveness of the Code following its initial 12-months implementation period. Should the results of this assessment not be satisfactory, the Commission may propose further initiatives, including of a regulatory nature.

3.Awareness raising and improving societal resilience

There is a need to join forces to counter disinformation and foster societal resilience. This can be done by empowering citizens and civil society and ensuring fact-based communication on the EU. Ahead of the European elections, the EU institutions have worked closely in the following areas:

·Raising awareness of disinformation: EU institutions launched a number of initiatives to raise public awareness of disinformation across EU Member States, including seminars, conferences and media briefings. Overall, hundreds of journalists participated, leading to an increased understanding and coverage of disinformation threats. Information sessions for visitors groups including during the European Youth Week in Brussels, hosted approximately 200 participants from civil society, universities and youth organisations. 

·Better communication on EU policies: To increase awareness and better understanding of the impact of the EU on their daily life, EU institutions provided proactive and multilingual communication to millions of EU citizens through social media 15 and through communication campaigns 16 . The European Parliament launched an awareness campaign called “This Time I'm Voting” 17 and several national authorities took similar actions. The Commission and the European External Action Service significantly expanded their positive communications through dedicated campaigns, reaching millions of citizens.

·Boosting the EU capacity to react to disinformation: The Commission worked together with the European External Action Service and the European Parliament to produce and distribute myth busting and awareness-raising materials, drawing on both academic and institutional expertise. The internal Commission Network against Disinformation operated regularly, in cooperation with other EU institutions, combining policy expertise and communication know-how. Several Commission Representations created pages on their websites rebutting longstanding myths about the EU and developed partnerships with fact checkers, editorial teams and other civil society initiatives to debunk disinformation.

·Strengthening societal resilience through media literacy: A high level of media literacy is key in enabling citizens to make informed decisions in the digital age. It is a pre-requisite for a vibrant, modern democracy. To scale up media literacy initiatives across the EU and highlight best practices, in particular at regional and national level, the Commission organised the first European Media Literacy Week with more than 320 events in the EU 18 . In line with the new audio-visual media services directive 19 , Member States will have to bolster media literacy measures in their national frameworks, in close cooperation with regulators and other stakeholders.

·Empowering civil society: The International Fact-Checking Network created a European branch of independent fact checkers covering 14 Member States and launched a website in 11 EU languages. The Commission will continue to facilitate the creation of a European multidisciplinary community of independent fact checkers and academic researchers. Under the Horizon 2020 programme, the Commission supported investments in new technologies for content verification and distribution through social media. In the Union’s neighbourhood, the Commission increased efforts to build links between EU and partner countries’ civil society. This allowed to identify and test community-driven solutions to fight disinformation.

4.Protecting the integrity of elections

Since the Facebook/Cambridge Analytica case and the revelations about election interference around the world, the Commission has acted decisively to protect the integrity of elections and increase societal resilience, in particular with its Election Package. These actions helped achieve concrete results in the following areas:

·Improved coordination of election authorities: The election networks consisting of authorities with competence relevant to electoral context, established in cooperation with Member States, led to increased preparedness for possible threats and better coordination of activities at national and EU level before the elections. Their activities included identification of threats and gaps, crisis preparedness, awareness-raising campaigns, efforts to counter disinformation, monitoring and enforcement of applicable rules and analyses of applicable national laws and procedures. The established collaboration effectively supported the preparation for elections in Member States and helped to address some of the differences in this complex area.

·Better protection against cyber threats: Practical exercises organised with the support of the European Union Agency for Network and Information Security (ENISA), the Commission and the Parliament helped strengthen preparedness and resilience towards cyber threats 20 . The EU institutions took steps to protect the social media accounts of the EU Institutions and of their selected key representatives 21 . With the adoption of a new legal framework 22 , the EU can now also impose sanctions (e.g. asset freeze, travel ban) to deter and respond to cyber-attacks that constitute an external threat to the Union or its Member States.

·Misuse of personal data: To clarify data protection obligations, the Commission issued guidance 23 on how to apply the General Data Protection Regulation in the electoral context. The European Data Protection Board also issued a statement on the use of personal data in political campaigns 24 committing to work with other relevant authorities to safeguard trust in the security and integrity of the elections. In addition, new measures introducing sanctions for infringements of data protection rules by European political parties entered into force 25 . The Commission recommended that Member States adopt the same approach at the national level.

·Boosting transparency: Member States, political parties, foundations and campaigns were invited to take measures to enhance the transparency of political communications and ads. The objective was to enable EU citizens to recognise online paid political ads and communications and identify who is behind them. For instance, this can be done by making information available on their websites relating to their expenditure on online activities and on targeting criteria used for ads.

Based on this experience, the Commission will adopt a comprehensive report in October 2019 on the elections, including on the implementation of the Elections package.

C.Conclusions

In the run-up to the European elections, the coordinated EU approach helped to ensure stronger preparedness and coordination in the fight against disinformation. The preliminary analysis shows that it contributed to expose disinformation attempts and to preserve the integrity of the elections 26 , while protecting freedom of expression. The highest turnout in the past twenty years (50.97 %) reflects the interest of the citizens for the Union and its importance for their lives.

However, there is no room for complacency. Disinformation is an evolving threat that requires continuous research to update our policy toolbox in line with new trends and practices. Malicious sources, both within and outside the EU, are constantly using new tactics, opting increasingly for smaller-scale local operations that are less likely to be detected and exposed. However, the objective remains the same: dividing our society and undermining the trust of citizens in democratic processes and institutions.

Protecting our democratic processes and institutions from disinformation and manipulation poses a long-term challenge and requires continuous efforts. This requires a joint effort of EU institutions and Member States and includes the need for adequate human and financial resources to better detect, analyse and expose disinformation campaigns and raising preparedness to address disinformation campaigns at EU and national level. The private sector, in particular online platforms, have a particular responsibility in tackling disinformation.

The Commission’s contribution to the informal EU27 leaders’ meeting in Sibiu on 9 May 27 made it clear: the rise of targeted disinformation campaigns will remain a major challenge for the future, and called for joint action by EU institutions and Member States to counter the threat.

The European External Action Service and the Commission, together with Member States, will further strengthen cooperation within the Rapid Alert System, including developing a common methodology for analysis and exposure of disinformation campaigns and stronger partnerships with international partners, such as G7 and the North Atlantic Treaty Organization.

Still this year, the Commission will report on the implementation of the Election Package and assess the effectiveness of the Code of Practice. On this basis further actions may be considered to ensure and improve our long-term response to the threat.

The Commission and the High Representative are committed to continue their joint efforts to protect our European democracy from disinformation and manipulation. It is our shared responsibility to safeguard EU citizens' right to factual, objective and reliable information.

(1)

     The Commission defines disinformation as ‘verifiably false or misleading information that is created, presented and disseminated for economic gain or to intentionally deceive the public, and may cause public harm. Disinformation does not include reporting errors, satire and parody, or clearly identified partisan news and commentary’(COM(2018)236). The aim of disinformation is to distract and divide, to plant seeds of doubt by distorting and falsifying facts, thus confusing people and weakening their faith in institutions and established political processes.

(2)

     JOIN(2018)36.

(3)

     The Election Package includes Commission Communication on securing free and fair European elections, COM (2018), 637; Commission Recommendation on election cooperation networks, online transparency, protection against cybersecurity incidents and fighting disinformation campaigns in the context of elections to the European Parliament, COM (2018) 5949; Commission Guidance on the application of Union data protection law in the electoral context, COM (2018) 638; Regulation (EU, Euratom) No 2019/493 of the European Parliament and of the Council of 25 March 2019 amending Regulation (EU, Euratom) No 1141/2014 as regards a verification procedure related to infringements of rules on the protection of personal data in the context of elections to the European Parliament, OJ L 85I , 27.3.2019, p. 7–10.

(4)

     The number of disinformation cases attributed to Russian sources and documented by the East Strategic Communication Task Force since January 2019 (998) doubled as compared to the same period in 2018 (434).

(5)

     See in particular Facebook announcements on the removal of pages, groups and accounts in other countries https://newsroom.fb.com/news/ .

(6)

     https://www.isdglobal.org/isd-publications/interim-briefing-propaganda-and-digital-campaigning-in-the-eu-elections/.

(7)

     For more on the disinformation cases on these issues, see EUvsDisinfo portal: https://euvsdisinfo.eu/disinformation-cases/.

(8)

     In May 2019 Microsoft became a signatory to the Code.

(9)

     A full report will be adopted by ERGA at the meeting of 20/21 June 2019.

(10)

     The Commission published these reports together with its assessment of the progress in the implementation of the Code of Practice.

(11)

     Such behaviour includes impersonation, spamming, misrepresentation of identity, links to ‘ad-farms’ and/or impostor websites. During the five-month period, Google took action against 131,621 EU-based ads accounts for violating its misrepresentation policies, and against 26,824 EU-based ads accounts for violating its policies on insufficient original content; also, it took action against 1,188 EU-based publisher accounts for violating its policies on valuable inventory. Facebook reported on some 1.2 million ads actioned in the EU for violating its policies on low quality or disruptive content, misleading or false content, or circumvention of its systems. Twitter reported rejecting 6,018 ads targeted at the EU for violation of its unacceptable business practices ads policy as well as 9,508 EU-targeted ads for violations of its quality ads policy.

(12)

     https://www.theguardian.com/world/2019/may/22/far-right-facebook-groups-spreading-hate-to-millions-in-europe.

(13)

     Avaaz (2019): Far Right Networks of Deception; Institute for Strategic Dialogue (2019): 2019 EU Elections Information Operations Analysis: Interim Briefing Paper; Avaaz & Institute for Strategic Dialogue (2019): Disrupted: Evidence of Widespread Digital Disruption of the 2019 European Parliament Elections.

(14)

     E.g. immigration, political values, civil and social rights, security and foreign policy, and environmental politics.

(15)

     The Commission regularly communicates via its social media accounts on the topic of disinformation: there were 57 posts from January 2018 to April 2019, across Twitter, Facebook, Instagram and LinkedIn, which generated 54 781 engagements and had an impact of 3 256 476 impressions (Twitter and Facebook only) and 385 655 reach (Facebook and Instagram only).

(16)

     A Europe that delivers (InvestEU, running in 16 Member states and reaching over 240 million potential contacts in 2017-2018), a Europe that empowers (EUandME, targeting around 100 million 18-35 year-olds in the EU, and reaching over 60 million potential contacts in 2018-2019) and a Europe that protects (targeting 35-55 year-olds – and reaching almost 60 million potential contacts in 2018-2019.)

(17)

     This campaign involved 300 000 supporters engaged in debating Europe in their local communities and social groups - both on and off-line. Over 340 three hundred pan-European organisations partnered up with Parliament to promote the elections.

(18)

     This initiative was carried out in cooperation with the Member States and took place from 18 to 22 March 2019.

(19)

     Directive (EU) 2018/1018 of the European Parliament and the Council of 14 November 2018.

(20)

     Two exercises organised by ENISA helped prepare national authorities for incidents aiming to create doubt about the legitimacy of the elections and for a scenario in which cyber-attacks were performed over critical infrastructures before and during European elections. Over 70 incident-handling experts from the EU Member States participated in the latest exercise in May 2019.

(21)

     Social Media Assurance Service (SMAS) provided by EU Computer Emergency Response Team.

(22)

     Council Regulation (EU) 2019/796 concerning restrictive measures against cyber-attacks threatening the Union or its Member States, OJ L 129I , 17.5.2019, p. 1–12; Council Decision (CFSP) 2019/797 concerning restrictive measures against cyber-attacks threatening the Union or its Member States, OJ L 129I , 17.5.2019, p. 13–19.

(23)

     COM(2018) 638.

(24)

     https://edpb.europa.eu/sites/edpb/files/files/file1/edpb-2019-03-13-statement-on-elections_en.pdf.

(25)

     Regulation 2019/493 amending Regulation 1141/2014 as regards a verification procedure related to infringements of rules on the protection of personal data in the context of elections to the European Parliament, OJ L 85I , 27.3.2019, p. 7–10.

(26)

     This has been recognised by independent actors and media as well. A study by the Oxford Internet Institute found that less than 4% of news sources shared on Twitter ahead of the European elections was disinformation content, while mainstream professional news outlets received 34% of shares. According to FactCheckEU , there was less disinformation than expected in the run up to the European elections and it did not dominate the conversation as it did around the past elections in Brazil, the United Kingdom, France or the United States.

(27)

     COM(2019)218.