Abstract: i social media, specie con l’avvento dell’intelligenza artificiale, potrebbero essere un validissimo strumento di un corretto veicolo informativo durante le campagne elettorali per le elezioni, specialmente quelle europee. Ma la prassi sconfessa molto spesso questa impostazione, poiché gli strumenti digitali vengono utilizzati per distorcere le informazioni nelle campagne elettorali, per diffondere fake news su partiti e candidati, per veicolare, più in generale, una disinformazione dilagante. Questo scritto si propone di analizzare questo fenomeno.
di Pierluigi Mascaro*
Abstract: i social media, specie con l’avvento dell’intelligenza artificiale, potrebbero essere un validissimo strumento di un corretto veicolo informativo durante le campagne elettorali per le elezioni, specialmente quelle europee. Ma la prassi sconfessa molto spesso questa impostazione, poiché gli strumenti digitali vengono utilizzati per distorcere le informazioni nelle campagne elettorali, per diffondere fake news su partiti e candidati, per veicolare, più in generale, una disinformazione dilagante. Questo scritto si propone di analizzare questo fenomeno.
Abstract: social media, especially with the advent of artificial intelligence, could be a very valuable tool for a proper information vehicle during election campaigns, especially European ones. But practice very often defeats this approach, as digital tools are used to distort information in election campaigns, to spread fake news about parties and candidates, and to convey, more generally, rampant disinformation. This paper aims to analyse this phenomenon.
Summary: 1. Introduction – 2. Artificial Intelligence, Social Media and Democracy – 3. Artificial Intelligence and online disinformation in European Parliament elections – 4. The European response to disinformation on social media – 5. Conclusions
1. Introduction
In the last decades, digital innovation and Artificial Intelligence (AI) are radically transforming democratic decision-making. In fact, policymakers and academia soon realized that the intensification of both the use of social media and the spread of online information could potentially be an underlying threat to the future of democracy[1].
Due to the lack of objective standards and the increasing number of deceptive content, social media have proved to be capable of negatively affecting the process of election through the distribution of content that can sway, deceive, or even manipulate voters.
Not surprisingly, disinformation figured at the center of the European agenda ahead of the 2019 European Parliament elections[2]: to meet these challenges, the EU employed a dynamic and multifaceted response that focused on data protection, transparency, cooperation with the private sector, promotion of diversity and credibility of information, rise of awareness, and the empowerment of the research community[3].
This paper starts with an examination of the academic debate regarding disinformation and the role of social media in this phenomenon. Then, it focuses on how the European institutions approached the problem, the steps that were taken to protect European citizens, and the process of elections. Ultimately, the paper concludes with an estimation of the EU’s overall approach and general observations regarding the future of digital disinformation.
2. Artificial Intelligence, Social Media and Democracy
It is very important to define what exactly Artificial Intelligence is and how it presents itself in elections: in particular, it is an optimal tool for election campaigns as its application into virtual platforms helps candidates and voters match with each other, or, if used malevolently, expose voters to fake news and disinformation to sway their vote to the opposing side[4].
Like most technological developments, AI maintains a range of benefits, limitations, and disadvantages. Its current use predominantly entails enhanced security, internet surveillance, social media marketing, and data analysis[5]. While all of these could be used for the “greater good” of society, there is an undeniable possibility of malevolent usage[6].
3. Artificial Intelligence and online disinformation in European Parliament elections
The 2019 European Parliamentary Elections posed an excellent example of how AI can have negative and positive implications on the electoral process: in fact, AI technologies have been increasingly utilized negatively to deter the democratic process through the distribution and promotion of disinformation on social media.
Capitalizing thus on mass audiences’ easy access to social media, especially Twitter, AI algorithms have been employed to misinform the public rapidly. Viral tweets spreading fake news, although easy to debunk with given time, can inflict great damage to elections and make them highly susceptible to short-term manipulation[7]. These fake news campaigns usually entailed extreme views antagonizing the institutions of the EU, thus undermining the people’s trust in their government and creating further problems for the European electoral process[8].
The first European initiatives to tackle this issue started in 2015 when the European Council recognized the threat in its conclusion on 20 March[9]; it is also important to notice that apart from the European instruments, each Member State has also developed similar initiatives on a national level.
Concerning the attention-based business model used by all free online media services, it relies on the promotion of specific content and advertisements to smaller groups of users based on their interests or individual characteristics[10]. Hence, this model contributes to a polarized and fragmented online information space that aids the dispensation of fake news and disinformation.
4. The European response to disinformation on social media
Given the plethora of threats and windows of opportunities, the EU adopted a multi-faceted response that incorporated legal (state legislation, self-regulation by the social network providers, and core regulation), technological (use of artificial intelligence), and social means (fact-checking, media literacy, and supporting quality journalism) (European Commission, 2020a). As noted in the previous section, the EU had already adopted corresponding legislation and created operational instruments that focused on the fight against disinformation.
A year before the elections, the EU adopted a more focused approach to disinformation with the Code of Practice. What distinguishes the code from other similar endeavors is that it was the first time globally that industry giants, such as Facebook, Google, Twitter, Mozilla, and Microsoft, agreed voluntarily to self-regulatory standards to fight disinformation (European Commission, 2018a). Following a definition of disinformation, the Code identifies five clusters of action to address the threat: a) scrutiny of ad placements (e.g., avoiding the monetization of disinformation-spreading websites); b) political advertising and issue-based advertising (e.g., more transparent distribution of political advertising); c) integrity of services (e.g., policies regarding the identity of the users); d) empowering consumers (e.g., investment in technological instruments prioritizing authentic information; e) empowering the research community (e.g., support of independent projects to track disinformation).
Later that year, the European Commission also published its Action Plan Against Disinformation[11]. Specifically, the Communication aimed to assist the task of securing free and fair democratic processes considering the size of the vote. In line with the Code, following the recognition of the threat of disinformation, the Commission set the four pillars which are the bases of its response: a) improving the capabilities of Union institutions to detect and expose disinformation; b) strengthening coordinated and joint responses to disinformation; c) mobilizing private sector to tackle disinformation; d) raising awareness and improving societal resilience.
Apart from the pillars, it also suggested maintaining the mandate of the East StratCom Task Force and revising the mandates of the Western Balkans and South Task Force. The Action Plan also introduced a new instrument, the Rapid Alert System that makes it possible to share data and give a coordinate response in Europe against disinformation campaigns. Specifically, the System is intended to link existing monitoring instruments inside and outside of the EU like the Emergency Response Coordination Centre and the EEAS Situation Room, along with the G7 Rapid Response Mechanism and the North Atlantic Treaty Organization. Due to its features, the Rapid Alert System is one of the major weapons of the EU against the threat of disinformation not only on the regional but also on the international level.
5. Conclusions
In the case of the European Union, the problem of malevolent AI usage was even greater as it had to endure all of the listed items and more. The EU saw an exponential increase in foreign hacking attempts, the mass spreading of fake news through social media, and election manipulation: to meet these challenges, the EU employed a dynamic and multifaceted response that focused on data protection, transparency, cooperation with the private sector, promotion of diversity and credibility of information, the rise of awareness, and the empowerment of the research community (Nenadic, 2019). Putting it simply, the EU incorporated legal (state legislation, self-regulation by the social network providers, and core regulation), technological (use of artificial intelligence), and social means (factchecking, media literacy, and supporting quality journalism) (European Commission, 2020a). Specifically, in a united effort to protect the legitimacy and authority of the EU, there was a vast array of initiatives launched that now serve as defenders against the malevolent use of AI.
Applications of AI are increasingly being used in daily life and as a result of the digital age, they will continue to dominate in the future. Given that, the EU response offers a map for tackling the issue of disinformation across the web that AI has powered. Furthermore, there was also an initiative in Brussels that was designed to track and halt hacking attacks that were coming from foreign nations and stop foreign nations from meddling in EU elections. Ultimately, the EU has not remained stagnant, but it has since then adopted two major policies: the 2020-2021 European Democracy Action Plan and the Digital Services Act[12].
It is now to be seen whether the electoral dysfunctions due to disinformation via social media will recur in the next European Parliamen elections in 2024 and in what form and, above all, whether the measures and precautions taken by the European institutions and member states to counter disinformation via social media are sufficient and adequately implemented.
If not, they will most likely be reviewed and corroborated in view of the next European Parliament elections.
* Dottore magistrale in Giurisprudenza.
[1] D. Kreiss, The Problem of Citizens: E-Democracy for Actually Existing Democracy, in Social Media Society, 1(2), 2015.
[2] European Commission, Action Plan on disinformation: Commission contribution to the European Council, 13-14 December 2018.
[3] I. Nenadić, Unpacking the “European approach” to tackling challenges of disinformation and political manipulation, in Internet Policy Review, 8(4), 2019.
[4] K. F. Ashoka, AI, Solidarity and The Future of Humanity, in Forbes, 2019.
[5] S. Grottola, Artificial intelligence and diplomacy: A new tool for diplomats?, in Diplomacy, 2018.
[6] I. P. Efthymiou-Egleton, T. W. Efthymiou-Egleton, & S. Sidiropoulos, Artificial Intelligence (AI) in Politics: Should Political AI be Controlled?, in International Journal of Innovative Science and Research Technology, 5(2), 2020, 49-51.
[7] J. Villasenor, How to deal with AI-enabled disinformation, in Brookings, 2020.
[8] M. Scott & L. Cerulus, Russian groups targeted EU election with fake news, says European Commission, in Politico, 2019.
[9] European Council, European Council meeting, 19 and 20 March 2015, conclusions.
[10] S. Joseph, Why the business model of social media giants like Facebook is incompatible with human rights, in The Conversation, 2018.
[11] European Commission, Code of Practice on Disinformation, 2018.
[12] European Commission, Coronavirus: EU strengthens action to tackle disinformation, 2020.
Summary: 1. Introduction – 2. Artificial Intelligence, Social Media and Democracy – 3. Artificial Intelligence and online disinformation in European Parliament elections – 4. The European response to disinformation on social media – 5. Conclusions
1. Introduction
In the last decades, digital innovation and Artificial Intelligence (AI) are radically transforming democratic decision-making. In fact, policymakers and academia soon realized that the intensification of both the use of social media and the spread of online information could potentially be an underlying threat to the future of democracy.
Due to the lack of objective standards and the increasing number of deceptive content, social media have proved to be capable of negatively affecting the process of election through the distribution of content that can sway, deceive, or even manipulate voters.
Not surprisingly, disinformation figured at the center of the European agenda ahead of the 2019 European Parliament elections: to meet these challenges, the EU employed a dynamic and multifaceted response that focused on data protection, transparency, cooperation with the private sector, promotion of diversity and credibility of information, rise of awareness, and the empowerment of the research community.
This paper starts with an examination of the academic debate regarding disinformation and the role of social media in this phenomenon. Then, it focuses on how the European institutions approached the problem, the steps that were taken to protect European citizens, and the process of elections. Ultimately, the paper concludes with an estimation of the EU’s overall approach and general observations regarding the future of digital disinformation.
2. Artificial Intelligence, Social Media and Democracy
It is very important to define what exactly Artificial Intelligence is and how it presents itself in elections: in particular, it is an optimal tool for election campaigns as its application into virtual platforms helps candidates and voters match with each other, or, if used malevolently, expose voters to fake news and disinformation to sway their vote to the opposing side.
Like most technological developments, AI maintains a range of benefits, limitations, and disadvantages. Its current use predominantly entails enhanced security, internet surveillance, social media marketing, and data analysis. While all of these could be used for the “greater good” of society, there is an undeniable possibility of malevolent usage.
3. Artificial Intelligence and online disinformation in European Parliament elections
The 2019 European Parliamentary Elections posed an excellent example of how AI can have negative and positive implications on the electoral process: in fact, AI technologies have been increasingly utilized negatively to deter the democratic process through the distribution and promotion of disinformation on social media.
Capitalizing thus on mass audiences’ easy access to social media, especially Twitter, AI algorithms have been employed to misinform the public rapidly. Viral tweets spreading fake news, although easy to debunk with given time, can inflict great damage to elections and make them highly susceptible to short-term manipulation. These fake news campaigns usually entailed extreme views antagonizing the institutions of the EU, thus undermining the people’s trust in their government and creating further problems for the European electoral process.
The first European initiatives to tackle this issue started in 2015 when the European Council recognized the threat in its conclusion on 20 March; it is also important to notice that apart from the European instruments, each Member State has also developed similar initiatives on a national level.
Concerning the attention-based business model used by all free online media services, it relies on the promotion of specific content and advertisements to smaller groups of users based on their interests or individual characteristics. Hence, this model contributes to a polarized and fragmented online information space that aids the dispensation of fake news and disinformation.
4. The European response to disinformation on social media
Given the plethora of threats and windows of opportunities, the EU adopted a multi-faceted response that incorporated legal (state legislation, self-regulation by the social network providers, and core regulation), technological (use of artificial intelligence), and social means (fact-checking, media literacy, and supporting quality journalism) (European Commission, 2020a). As noted in the previous section, the EU had already adopted corresponding legislation and created operational instruments that focused on the fight against disinformation.
A year before the elections, the EU adopted a more focused approach to disinformation with the Code of Practice. What distinguishes the code from other similar endeavors is that it was the first time globally that industry giants, such as Facebook, Google, Twitter, Mozilla, and Microsoft, agreed voluntarily to self-regulatory standards to fight disinformation (European Commission, 2018a). Following a definition of disinformation, the Code identifies five clusters of action to address the threat: a) scrutiny of ad placements (e.g., avoiding the monetization of disinformation-spreading websites); b) political advertising and issue-based advertising (e.g., more transparent distribution of political advertising); c) integrity of services (e.g., policies regarding the identity of the users); d) empowering consumers (e.g., investment in technological instruments prioritizing authentic information; e) empowering the research community (e.g., support of independent projects to track disinformation).
Later that year, the European Commission also published its Action Plan Against Disinformation. Specifically, the Communication aimed to assist the task of securing free and fair democratic processes considering the size of the vote. In line with the Code, following the recognition of the threat of disinformation, the Commission set the four pillars which are the bases of its response: a) improving the capabilities of Union institutions to detect and expose disinformation; b) strengthening coordinated and joint responses to disinformation; c) mobilizing private sector to tackle disinformation; d) raising awareness and improving societal resilience.
Apart from the pillars, it also suggested maintaining the mandate of the East StratCom Task Force and revising the mandates of the Western Balkans and South Task Force. The Action Plan also introduced a new instrument, the Rapid Alert System that makes it possible to share data and give a coordinate response in Europe against disinformation campaigns. Specifically, the System is intended to link existing monitoring instruments inside and outside of the EU like the Emergency Response Coordination Centre and the EEAS Situation Room, along with the G7 Rapid Response Mechanism and the North Atlantic Treaty Organization. Due to its features, the Rapid Alert System is one of the major weapons of the EU against the threat of disinformation not only on the regional but also on the international level.
5. Conclusions
In the case of the European Union, the problem of malevolent AI usage was even greater as it had to endure all of the listed items and more. The EU saw an exponential increase in foreign hacking attempts, the mass spreading of fake news through social media, and election manipulation: to meet these challenges, the EU employed a dynamic and multifaceted response that focused on data protection, transparency, cooperation with the private sector, promotion of diversity and credibility of information, the rise of awareness, and the empowerment of the research community (Nenadic, 2019). Putting it simply, the EU incorporated legal (state legislation, self-regulation by the social network providers, and core regulation), technological (use of artificial intelligence), and social means (factchecking, media literacy, and supporting quality journalism) (European Commission, 2020a). Specifically, in a united effort to protect the legitimacy and authority of the EU, there was a vast array of initiatives launched that now serve as defenders against the malevolent use of AI.
Applications of AI are increasingly being used in daily life and as a result of the digital age, they will continue to dominate in the future. Given that, the EU response offers a map for tackling the issue of disinformation across the web that AI has powered. Furthermore, there was also an initiative in Brussels that was designed to track and halt hacking attacks that were coming from foreign nations and stop foreign nations from meddling in EU elections. Ultimately, the EU has not remained stagnant, but it has since then adopted two major policies: the 2020-2021 European Democracy Action Plan and the Digital Services Act.
It is now to be seen whether the electoral dysfunctions due to disinformation via social media will recur in the next European Parliamen elections in 2024 and in what form and, above all, whether the measures and precautions taken by the European institutions and member states to counter disinformation via social media are sufficient and adequately implemented.
If not, they will most likely be reviewed and corroborated in view of the next European Parliament elections.
