Facilitating Collective Action in Agentic IS Platforms: The Case of Decentralized Autonomous Organizations

Abstract

An increasing number of platform organizations are run by agentic algorithms. While much is known about algorithmic management on centralized platform organizations such as Uber, where human agency is limited through data-driven surveillance and control, decentralized organizing has begun to emerge in agentic IS platforms, characterized by conjoined human and algorithmic agency. We examine this shift toward decentralized organizing by analyzing multiple cases of decentralized autonomous organizations (DAOs) that leverage blockchain technology and seek to expand, not limit, collective human action while being run by agentic algorithms. The phenomenon of DAOs gives rise to a puzzle: How is collective action facilitated in agentic IS platforms when they are increasingly run by algorithms? Our findings show that agentic algorithms have the ability to facilitate collective action that is aligned around a common human purpose dynamically negotiated through adaptation of the set of algorithms running the organization. We explain how collective action is enacted in DAOs, as an example of agentic IS platforms, presenting a grounded model developed inductively based on our multiple-case study. Our study complements and extends the work of Baird and Maruping (2021) by expanding the focus of analysis of agentic IS artifacts to agentic IS platforms and considering facilitation as well as delegation. Our detailed findings about decentralized algorithmic management and decentralized management of algorithms in DAOs extend the literature on platform organizing, which thus far has paid only limited attention to the unique mechanisms at play in the context of conjoined agency.