AI in Business and Society

Our track is concerned with research on the design and evaluation of sociotechnical AI- based systems that achieve multi-sided outcomes and are meaningful to the businesses and/or society as well as the use and consequences of sociotechnical AI systems. Particularly, we welcome studies that examine impactful “AI design principles” and research that yields new principles for the design and evaluation of AI tools. This also includes studies that take on a hybridization approach and present designs of interesting human-AI hybrids in different contexts. We seek studies that are theoretical, empirical, or, technical, quantitative or qualitative, as long as the research yields impactful new insights. By “impactful,” we mean that the design principles should pivot from the existing research, theory, and practice; specifically, authors will need to demonstrate through their understanding of the existing opus, how their research builds on current knowledge on the topic, and not simply state that fact. By “multi-sided outcomes,” we mean an effective AI-based tool or system which achieves value not just for the developer or the corporation using it on consumers, employees or contributors but for those other users as well. By “socially meaningful multi-sided outcomes,” we mean that the value to the users should be in ways that go far beyond simple recommendations for movie choices, but improve people’s lives in fundamental ways (e.g., closing the income inequality gap, dampening systemic racial biases, reducing information silos, engaging with the challenges of global warming, improving the safety of seniors’ homes, etc.). By “sociotechnical AI,” we mean that studies should seek to open and provide insights into the black box of the user, the ecosystem of use and development, and the technology around it. For example, discussions of the AI tools should include the data and algorithms they are built on, how users are enticed into engaging with these tools, and what is the ecosystem that institutionalizes the tool-usage patterns that harm or foster the multi-sided outcomes. Through our interest in human-AI hybrids, we are also seeking studies that not only focus on the design of AI-based tools for the user but shed more light into how humans and AI can work collaboratively on a task.

Track Chairs:
Ann Majchrzak, University of Southern California
Kevin Hong, University of Houston
Saonee Sarker, University of Virginia