Recent developments show that electoral malpractice is a growing problem, not only in authoritarian and hybrid regimes, but also in democracies. Therefore, election monitoring is an indispensable element for safeguarding electoral integrity.[1] Until recently, monitoring activities were primarily conducted by international and domestic organisations in traditional ways, such as monitoring polling stations, parallel vote counting, cross-checking the validation of polling station vote counts and reporting electoral malpractice. However, these methods are not able to detect computational propaganda[2] and foreign influence via social media, as in the case of Russian interference in the 2016 US elections and the 2016 UK Brexit referendum. In line with recent developments, monitoring organisations have begun to adopt new methods to monitor social media. However, they are still in the early stages of finding the most appropriate methodological approaches and tools.
Although the use of disinformation and efforts to disrupt elections have a long history, since 2016 social media has become the primary domain that both state and non-state actors have actively used to interfere in elections. Therefore, governments have emerged as a new actor in election monitoring by using AI tools to monitor social media in order to ensure electoral integrity. Many EU countries, including Germany, Sweden, France, Denmark and the Czech Republic, have begun to use AI systems for social media monitoring during elections (Schmuziger Goldzweig et al., 2019, pp. 15–16). In addition, the EU has been working towards taking protective measures, from regulating political advertising to countering cyber attacks, in order to conduct free and fair elections as per the European Democracy Action Plan, which is being implemented gradually until 2023 (Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions on the European Democracy Action Plan, 2020). In this regard, social media monitoring has been included in the EU Election Observation Missions’ duties, and the East StratCom Task Force has been focusing on activities undertaken by Russia in an attempt to disrupt elections.
These developments show that social media monitoring has become a central domain in election monitoring, in order to support free and fair elections and uphold the quality of democracy. However, while many studies explore how both state and non-state actors disseminate disinformation and interfere in elections via social media, research has yet to systematically investigate how governments have responded. To date, little is known about how they have faced these challenges and found solutions with the help of AI systems. To address this gap, this research aims to address the challenges and opportunities created by using AI tools on social media in order to support government monitoring and help them to ensure electoral integrity. Furthermore, it will consider the ethical risks and social implications of using AI tools and suggest possible improvements. Since the EU is the region where new regulations have been brought out, and a few EU countries have started to utilise AI systems for social media monitoring, they offer a good starting point to explore and discuss the developments. For this purpose, European countries will be studied from 2016 to 2021.
References
Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions on the European Democracy Action Plan, Pub. L. No. COM/2020/790 (2020).
Norris, P. (2013). The New Research Agenda Studying Electoral Integrity. Electoral Studies, 32(4), 563–575. https://doi.org/10.1016/j.electstud.2013.07.015
Schmuziger Goldzweig, R., Lupion, B., & Meyer-Resende, M. (2019). Experiences of Social Media Monitoring During Elections Cases and Best Practice to Inform Electoral Observations Missions. Open Society Foundations.
Woolley, S. C., & Howard, P. N. (2017). Computational Propaganda Worldwide: Executive Summary (Working Paper 2017.11; Working Paper 2017).
[1] The term “electoral integrity” is used in its broadest sense to refer to an attachment to international standards and global norms during the electoral period (Norris, 2013, p. 564).
[2] Computational propaganda is defined as “the use of algorithms, automation, and human curation to purposefully distribute misleading information over social media networks” (Woolley & Howard, 2017, p. 3).
Governments’ use of AI systems in social media monitoring during elections 