Project

Back to overview

AI and Democracy: an Experiment

English title AI and Democracy: an Experiment
Applicant Duberry Jérôme
Number 190509
Funding scheme Spark
Research institution Global Studies Institute Université de Genève
Institution of higher education University of Geneva - GE
Main discipline Political science
Start/End 01.02.2020 - 31.08.2021
Approved amount 113'574.00
Show all

All Disciplines (2)

Discipline
Political science
Information Technology

Keywords (8)

Democracy; Information and communication technologies; Artificial intelligence; Digital technologies; ICT; Political participaton; Citizen participation; AI

Lay Summary (French)

Lead
Le concept de démocratie augmentée (Augmented Democracy), développé par Hidalgo et son équipe (2018) au Massachusetts Institute of Technology (MIT), propose d'explorer comment la l'intelligence artificielle pourrait soutenir des processus décisionnels démocratiques. Il repose sur l'hypothèse que (1) plus d'information et plus d'instances de participation ne sont pas nécessairement la solution au déclin de la participation citoyenne dans les États démocratiques libéraux, et que (2) l'IA s'infiltre progressivement dans tous les aspects de nos vies, y compris la politique. C'est précisément ce que ce projet interdisciplinaire vise à examiner : comment l'IA pourrait favoriser la démocratie, la participation citoyenne et soutenir les décideurs politiques.
Lay summary
Après plus de 15 ans de développement, le Conseil fédéral a décidé de suspendre temporairement la mise en oeuvre du vote électronique en Suisse. Cette décision illustre bien la difficulté de définir le rôle des technologies numériques dans les démocraties libérales. D'une part, la baisse de la participation politique exige de nouveaux instruments pour favoriser l'engagement des citoyens. D'autre part, les préoccupations croissantes en matière de cybersécurité et les campagnes de désinformation obligent les gouvernements à réduire les initiatives de vote électronique. Dans le débat sur le rôle des technologies numériques dans les démocraties libérales, l'intelligence artificielle (IA) est probablement la plus controversée. Cette technologie suscite les plus grands espoirs (par exemple, l'IA peut aider à résoudre des problèmes mondiaux tels que le changement climatique) et soulève en même temps de nombreuses préoccupations (par exemple, l'IA finira-t-elle par prendre en charge les négociations politiques et prendre toutes les décisions ?). Si l'avenir de cette technologie n'est pas encore clair, il est néanmoins crucial de proposer quelques orientations qui puissent bénéficier aux citoyens et aux démocraties libérales en général.  

Puisque l'avenir des démocraties est remis en question par certains universitaires, dirigeants politiques et entrepreneurs technologiques, ce projet souhaite explorer si une application d'IA, basée sur le concept de "Augmented Democracy" peut favoriser la démocratie libérale. Ce projet de recherche souhaite acquérir, en un an, suffisamment de données pour conclure si cette application d'IA mérite d'être explorée plus avant. Cette expérience devrait donc nous permettre de savoir si les "digital twins" peuvent servir de base à des projets de recherche plus ambitieux, en particulier pour étudier leur potentielle résistance aux campagnes de désinformation et aux cyber-attaques. Les données recueillies et le rapport devraient également contribuer à un débat public sur l'avenir de la démocratie et le rôle d'AI et des TIC en général.
Direct link to Lay Summary Last update: 22.01.2020

Lay Summary (English)

Lead
The concept of Augmented Democracy (AD), developed by Hidalgo and his team (2018) at the Massachusetts Institute of Technology (MIT) proposes to explore how AI could foster democracy. It is based on the assumptions that (1) more information and more instances of participation are not necessarily the solution to the decline of citizen participation in liberal democratic states, and that (2) AI is progressively infiltrating all aspects of our lives, including politics. This is precisely what this interdisciplinary project aims to examine: how AI could foster democracy, citizen participation and support policy-makers.
Lay summary

After over 15 years of development, the Federal Council decided to temporarily put on hold the implementation of the electronic vote in Switzerland. This decision illustrates well the difficulty to define the role of digital technologies in liberal democracies. On one hand, the declining rates of political participation calls for new instruments to foster citizen engagement. On the other hand, rising cybersecurity concerns and disinformation campaigns force governments to scale down e-voting initiatives. When discussing the role of digital technologies in liberal democracies, artificial intelligence (AI) is probably the most controversial. This technology triggers the greatest hopes (e.g. AI can help tackle global issues such as climate change) and at the same time raises many concerns (e.g. will AI eventually take over political negotiations and make all decisions?). If it is not clear yet what the future of this technology will be, it is nevertheless crucial to propose some orientations that can benefit citizens and liberal democracies at large.  

Since the future of democracies is questioned by some scholars, political leaders and tech entrepreneurs, this project wishes to explore if an AI application can foster liberal democracy. This research project wishes to acquire, in one year, enough data to conclude if an AI application based on the Augmented Democracy concept is worth exploring further. Hence, this experiment should inform us if digital twins can become the basis for larger research projects (with more participants, and multi-level policy-making processes), which would also explore their resilience to disinformation campaigns and cyberattacks (the two main reasons why the Federal Council decided to scale down e-voting in Switzerland). The data collected and the report should also enable policy-makers and civil society to contribute to a public debate about the future of democracy and the role of AI and ICT at large.

 
Direct link to Lay Summary Last update: 22.01.2020

Responsible applicant and co-applicants

Employees

Abstract

After over 15 years of development, the Federal Council decided to temporarily put on hold the implementation of the electronic vote in Switzerland. This decision illustrates well the difficulty to define the role of digital technologies in liberal democracies. On one hand, the declining rates of political participation calls for new instruments to foster citizen engagement. On the other hand, rising cybersecurity concerns and disinformation campaigns force governments to scale down e-voting initiatives. When discussing the role of digital technologies in liberal democracies, artificial intelligence (AI) is probably the most controversial. This technology triggers the greatest hopes (e.g. AI can help tackle global issues such as climate change) and at the same time raises many concerns (e.g. will AI eventually take over political negotiations and make all decisions?). If it is not clear yet what the future of this technology will be, it is nevertheless crucial to propose some orientations that can benefit citizens and liberal democracies at large. The concept of Augmented Democracy (AD), developed by Hidalgo and his team (2018) at the Massachusetts of Technology (MIT) proposes to explore how AI could foster democracy. It is based on the assumptions that (1) more information and more instances of participation are not necessarily the solution to the decline of citizen participation in liberal democratic states, and that (2) AI is progressively infiltrating all aspects of our lives, including politics. This is precisely what this interdisciplinary project aims to examine: how AI could foster democracy, citizen participation and support policy-makers.
-