PROJECT SUMMARY
Assessing Prospective Harms (vs Benefits) associated with ADM
Focus Areas: News and Media, Transport and Mobility, Health, and Social Services
Research Programs: Data, Machines, Institutions
Status: Completed
The project (which is now completed) was set up as a preliminary exercise in assessing prospective harms vs prospective benefits associated with ADM as a first step to amelioration. It took a two-pronged approach: firstly, focusing on individual and social harms/costs that may be associated with automated or semi-automated data processing (including collection, retention, dissemination, and other uses of data) – versus prospective benefits; and secondly, assessing the levels of risk of these harms ranging from nebulous to very significant (and acknowledging there may be uncertain outcomes and uneven distributions). The overall aim was thus to have a fuller appreciation of harms and risks as a precursor to thinking practically about amelioration/mitigation of costs.
More specifically, the project was geared to questions of elaborating and understanding the range of prospective harms associated with loss of control over data processing for individuals, groups and society, and indeed the entirety of the living world, as a first step to finding solutions such as changes in law, or social practices, or business methods, or technologies (or some combination of these).
The principal activity of the project was to have a series of workshops planned, organised and hosted by the coordinators CI Richardson, AI Roberts and Postdoc Jiménez (with administrator Astari.Kusumawardani providing support). The workshops featured the work of diverse ADM+S CIs, AIs, Researchers and Affiliates and adopted an intense mode of interrogation and discussion along with suggestions. The aim was to assist ADMS personnel with the preparation of reports, books and scholarly articles (as well as share insights and ideas).
Topics and presenters in the workshop series included the following:
•March: Aitor Jiménez (Megan Richardson chair), Crimes of digital capitalism
•March: Ariadna Matamoros- Fernández, Rosalie Gillett, Anjalee de Silva (Aitor Jiménez chair), •Gendered harm
•April: José-Miguel Bello Villarino, Henry Fraser (Megan Richardson chair), Where residual risks reside: a comparative approach to AI risk management under the EU’s AI Act Proposal
•April: Jake Goldenfein (Megan Richardsonchair) How competing constructions of humans legitimize online advertising
•May: Simon Coghlan, Christine Parker (Andy Roberts, chair), A preliminary framework for understanding how ADM/AI technologies can harm non-human animals
•June: Lisa Archbold (Andy Roberts chair), Children’s developmental privacy
•July: Frank Pasquale/Jeannie Paterson (Megan Richardson chair: co-hosted with CAIDE), Automated grace: toward more humane benefits administration via AI
•August: James Meese (Megan Richardson chair), Regulating news recommendation: looking beyond harm
•September: Megan Richardson (Jeannie Paterson chair – co-hosted with CAIDE), Trust norms and data rights
•October: Ariadna Matamoros- Fernández, Louisa Bartolo, Luke Troynar (Aitor Jiménez chair), Addressing harmful humour as an online safety issue
•November: Damian Clifford (Megan Richardson chair), Data protection and (in)accuracy
PUBLICATIONS
Harm to Nonhuman Animals from AI: a Systematic Account and Framework, 2023
Parker, C., Coghlan, S.
Humour as an online safety issue: Exploring solutions to help platforms better address this form of expression, 2023
Matamoros-Fernández, A., Bartolo, L., Troynar, L.