Throwing keywords at the internet : emerging practices and challenges in human rights open source investigations




Banchik, Anna Veronica

Journal Title

Journal ISSN

Volume Title



Human rights researchers are increasingly turning to the internet to discover, collect, and preserve user-generated content (UGC) documenting human rights abuses. The proliferation of UGC and other kinds of “open source” (publicly accessible) information online make available more information than ever before about abuses. Only—UGC may be incapable of verification, buried online, or in peril of deletion by platform content moderation, state coordinated flagging campaigns, or users themselves. Some contexts might produce a scarcity of information while others generate a deluge of data, complicating already lengthy ad hoc verification procedures. Moreover, UGC collection, recirculation, and preservation may endanger or go against the wishes of witnesses and their families. Drawing on interviews with open source practitioners and experts as well as a year of ethnographic fieldwork at the Human Rights Investigations Lab at U.C. Berkeley’s Human Rights Center, this dissertation examines the practices, logics, and narratives by which a growing number of researchers are applying open source investigative techniques to human rights advocacy and fact-finding. First, amidst increased scrutiny of platforms content moderation, this study highlights the more nuanced roles of platform design elements including algorithms and search functionality in shaping both UGC’s discoverability and investigators’ workarounds—underscoring the ephemerality of UGC itself and of the informational infrastructures through which UGC is sought. Second, this dissertation offers a typology to synthesize how an array of broader social and structural factors impinge on the verifiability of UGC from varied conflicts and contexts, building on prior research pointing to factors impacting the volume of content available about a given conflict as well as scholarship suggesting that the inclusion of “verification subsidies” (McPherson 2015a) into photos or video footage heightens content’s verifiability. Third, this study examines the emergence of human rights advocates and researchers as self-ascribed content stewards and safeguards. In addition to describing investigators’ affective and ethical commitments to UGC, I point to how pervasive “consent-cutting” discourses combine with decisions to refrain from contacting uploaders in ways that effectively normalize the sustenance of communication and consent gaps with content uploaders, raising ethical questions about responsible data collection and usage.



LCSH Subject Headings