Artificial Intelligence, Democracy and Human Dignity

External project members:

Anna-Sara Lind, Professor of Public Law at the Department of Law, Uppsala University (project leader)
Johan Wejryd, PhD in Political Science, researcher at IBF, Uppsala University
Oliver Li, PhD in Philosophy of Religion, researcher at the Center for Multidisciplinary Research on Religion and Society (CRS), Uppsala University
Johan Eddebo, PhD, associate professor in Philosophy of Religion, researcher at the Center for Multidisciplinary Research on Religion and Society (CRS), Uppsala University
Lars Karlander, PhD in Constitutional Law, researcher at the Center for Multidisciplinary Research on Religion and Society, Uppsala University.

Background

As artificial intelligence is growing in scope and strength, it raises the issue of whether it could develop independence. A research project at Uppsala University led by Anna-Sara Lind will monitor this development and how it impacts on society.

One of the fundamental issues for the researchers to investigate is whether strong AI is even possible and, if so, how it would affect democracy, human dignity, human rights and other ethical questions. In the historical paradigm shift entailed by today’s technological development, there is a risk of humans having less influence in the world, instead of more.

This gives rise to a need for a deeper understanding of the values created and transferred as technology based on artificial intelligence becomes increasingly common. The research project is intended to create a clearer idea of which effects artificial intelligence may have in terms of rights, ethics, world views and social institutions at a global level.

Overall purpuse:

On a foundation of philosophy, political science, and jurisprudence this project will investigate the metaphysical possibility as well as the democratic and legal implications of genuine artificial subjectivity in three interlinked lines of inquiry.

1. In the first line of inquiry (1) we aim to provide a thorough analysis of the metaphysical viability of strong AI in the sense of artificial phenomenal subjectivity, that is whether it is reasonable to assume that any artificial entity ever could possess consciousness and free will. This will be done mainly through a detailed examination of the philosophical soundness of the ontological preconditions of actual strong AI. Consequently, an account of the metaphysical implications of the affirmation of strong AI will be provided, i.e. an account of what in effect is being implicitly stated when one makes the popular assessment that strong AI is realizable. Additionally, we will examine the possibility of instantiating free will in an artificial subject in two steps. We will begin by integrating the possible conceptualizations of free-willed AI into the available ontological frameworks of subjectivity, and go on to determining the philosophical soundness of the resulting alternatives.
   
   
2. In a second line of inquiry (2) the value-related effects of the implicit ontological truth-claims of the affirmation of strong AI can then be analyzed. This will enable us to assess how affirming strong AI implicates on values and ideas central to contemporary liberal democracy, such as that of a human right to self-determination and the nature of moral agency. Implementing these analyses, the project will explore the effects of affirming strong AI with regard to democratic inclusion. The scholarly debate on democratic inclusion has hence been focused on enfranchisement conditions for particular human individuals. Given that advanced cognitive ability and the capacity of self-determination are the basic prerequisites for political enfranchisement, the ascription of these attributes to artificial entities enables a case for their inclusion in governance and political decision-making. The normative legitimacy of such inclusion will be tested against the yardsticks of contemporary normative democratic theory. Furthermore, the political viability of democratic inclusion of AI, in relation to popular attitudes (sociological legitimacy), will explored. The latter will be done by means of survey experiments regarding the perceived legitimacy of different decisions being taken by AI:s or not.
   
   
3. In the third line of inquiry (3) the project will scrutinize the concrete legal implications of the affirmation of strong AI. As with political enfranchisement, the extension of legal subjectivity to certain artificial intelligence can be seen to follow from the ascription of advanced cognitive ability to such entities. We will establish criteria that can be used to evaluate whether AI-entities could be fitted into the existing categories of legal subjects, relating to basic human rights and human dignity. If – as is likely – AI-entities will not readily suit into these categories, we will explore the possibilities for creating a new form of legal subjects.