Academic WorkshopMarch 14 2023

The Social Life of Algorithmic Harms

March 10–11, 2022

Emanuel Moss
Jacob Metcalf

On March 10–11, 2022 Data & Society held an online workshop that brought together researchers and advocates from around the world to consider novel algorithmic harms that are underappreciated by current approaches to AI governance, as well as methods that are emerging to better understand, evaluate, and assess those harms. Rather than start from the problems for which developers can most readily identify technical solutions — like privacy and unfairness, which have received the lion’s share of attention in the world of AI governance — the workshop began by looking at the social life of algorithmic harms that travel beyond the boundaries of the technical systems. In addition to interdisciplinary discussions and debates, participants provided feedback for fifteen papers-in-progress that each took a novel approach to the theme.

In his introduction to The Social Life of Things, Arjun Appadurai invites anthropologists to “follow the things themselves for their meanings are inscribed in their forms, their uses, their trajectories… [that manifest in] human transactions and calculations that enliven things.” Along similar lines, the workshop’s participants began by following algorithmic systems themselves to see how their meanings manifest in social exchanges and lived consequences. Among the myriad of such meanings, they identified poignant instances that offered a new window into algorithmic harms that have gone largely unnoticed by technologists and other AI professionals — harms inflicted at the intersection of medicine and technology, through applications in child protective services, and as impacts on the natural environment. In mapping the social life of algorithmic harms, workshop participants challenged the notion that algorithmic harms are merely technical problems in need of technical solutions. 

In 2023, we began publishing a series of essays that grew out of the workshop, written by participants, on Data & Society’s Points blog. By highlighting novel forms of algorithmic harms and how their implications change as they move through social exchanges, the series seeks to expand our vocabulary of these harms — and with it, our capacity to defend ourselves against them.

Acknowledgments and Credits

Hosts: Jacob Metcalf and Emanuel Moss

Producer: CJ Brody Landow

Session Facilitator: Nina Mehta

Workshop Production Assistant: Eli Eley

Platform Designer: Ryan Anderson

Keynote Presenters: Ali Alkhatib, Margot E. Kaminski  and Tawana Petty

Keynote Lightning Talks Illustrator: Zahra Zainal

Keynote Lightning Talks Video Editor: Collin Leitch


— Thank You —

This workshop is partially supported by the National Science Foundation, Award #1704425. It is the result of ongoing collaboration between Emanuel Moss and Jacob Metcalf.

Designing this workshop included brainstorming sessions and feedback on initial drafts of the call with Jenna Burrell, Sareeta Amrute, Ranjit Singh, Elizabeth Anne Watkins, Emnet Tafesse, Robyn Caplan, Siera Dissmore, and Ania Calderon. Additional Data & Society teams contributed their expertise to this call, especially CJ Brody Landow, Chris Redwood, Veronica Eghdami, Joanna Gould, and Sam Hinds. Production benefited from conversations with and resources from Zara Rahman, Mona Sloane, Nazelie Doghramadjian, Sarah Welsh, Noah Landow, and more. 

We also appreciate the following people and teams who helped shape our keynote video collection: Scheherazade Washington Parrish and Allied Media Projects, Data & Society’s Communications, Accounting, Events, and AIGI teams, as well as Tamara K. Nopper, Chris Gilliard, and Sareeta Amrute.