Designed agency in collaborations: Exploring cross-sector collaboration in Finland’s artificial intelligence programme AuroraAI.

Abstract

Prospects of advanced digitalisation, namely applications of algorithmic computation and artificial intelligence, are expected to improve data-driven decision-making in business and government alike. Overshadowed by this vast momentum in technology and predicted progress, societal questions of human dignity and democratic participation in anticipation of futures are fading out of attention. Cross-sector collaborations in the public sector are perceived as a viable means to address complex socio-technical problems, such as the above, as part of an emerging shift from market- and performance-focused governance and towards public good. Simultaneously, the discipline of design, increasingly permeating other fields, sees progressive application in the public realm where it provides encouraging means of participatory engagement to support the reorientation of governance towards the human being.

My research takes a critical perspective on the preliminary, pre-2020, preparations of AuroraAI, the Finnish national programme for artificial intelligence, and interconnected cross-sectoral service provision. By developing a human-centric lens of design, the mixed-methods study constructively investigates barriers in the collaborative development and how these closely relate to the currently present and omitted actors and their respective agency. Normative aspects inherent to questions of fair participation in the creation of public good and joint futures are substantiated with the reviewed literature ranging from design to political theory.

The thesis highlights the importance of actively nurturing intangible structures of trust and mutual understanding to establish ownership and equity in decision-making. Different levels of agency among actors in the programme appear to be profoundly determined by consciously and unconsciously taken design choices regarding the structures that create the foundations for the processes of collaborative engagement. If agency is the capacity of an actor to exert power in a given context, this capacity can be deliberately or unintentionally limited and expanded; hence agency is open to be designed towards a preferred level of capacity. In the context of collaborations, designed structures, rules and norms then become the main lever to manipulate agency and thereby power dynamics, according to prefigured values and principles. Thus far, the collective agency in AuroraAI seems to be affected by the ramifications of structural limitations regarding actor involvement, open communication and the collaborative engagement regarding a partly prefigured techno-centric agenda. I propose a strategic reframing towards jointly deliberated values of public good within a wider network of actors in their self-determination of digital futures. Structures that guarantee continuous public engagement are not only considered a matter of principle but as a direct means for sustaining relational trust between the government and civil society, as well as to augment internal goal consistency and enhanced legitimacy. Hence, the study acknowledges the design of agency via formal and informal structures to be the reflection and reproduction of value-decisions regarding power dynamics in a collaboration and its political environment.