AI, the Social Contract and Democracy

Technology and democracy transform in tandem. The impact of the printing press on protestant politics first and the French revolution later on or that of Cambridge Analytica on recent U.S. election outcomes suggests as much. The research area of ‘AI, the Social Contract and Democracy’ cuts to the heart of this phenomenon, while firmly anchoring it in the analytical paradigm of democracy. The growing spread of intelligent algorithms imply fundamental shifts in the social contract underlying our modern democracies and some have talked about a digitized and algorithmic social contract. The technology sector experiences that new technological possibilities rub up against rules and social categories made in a different epoch, politicians and regulators realize that they are often late and badly suited to foresee the challenges of AI, and citizens are left to trust processes they have very little knowledge about and influence over.

The social contract, which we have lived by in modern democracies, is getting de-facto reformulated. This means that we need to conceptually and collectively rethink and (perhaps) revise the agreement among citizens on how to form an organized society, assign responsibility and legitimize the authority of the state and other entities (like professions) over certain aspects of individual citizens’ lives to maintain social order and guarantee the respect of human rights we have agreed upon. Among other visible phenomena, intelligent algorithms have transformed how the “people” are constituted, persuaded and represented in our “rule by the people” (the literal meaning of democracy) and how the public deliberative space functions, moving towards amalgamation of reasoning, lobbying, marketing and manipulation in a way that cannot be understood in isolation from the technology enabling, shaping and leveraging it.

Within this wide research area, the reseach project investigates consequences and challenges of AI on specific aspects of the social contract in our modern democracies, develop an analytical framework for governing AI development, and inform the design of AI systems which uphold our democratic principles, like public deliberation, representation and universal human rights. While we see an increased interest in the impact of AI on democracy, it often lacks a granular understanding of how AI shapes the tensions between the individual and the common, the private and the public, the free and the governed. Results of such an investigation can contribute significantly to developing AI systems that are not only explainable and transparent, but also grounded in a social scientific and humanistic understanding of democratic processes at large.