Sneaking AI through the Back Door: Constructing the Identity of Capitol Hill Rioters through Social Media Images and Facial Recognition Technologies
Research output: Contribution to journal › Journal article › Research › peer-review
Standard
Sneaking AI through the Back Door : Constructing the Identity of Capitol Hill Rioters through Social Media Images and Facial Recognition Technologies. / Mortensen, Mette.
In: Information, Communication & Society, 2024, p. 1-17.Research output: Contribution to journal › Journal article › Research › peer-review
Harvard
APA
Vancouver
Author
Bibtex
}
RIS
TY - JOUR
T1 - Sneaking AI through the Back Door
T2 - Constructing the Identity of Capitol Hill Rioters through Social Media Images and Facial Recognition Technologies
AU - Mortensen, Mette
PY - 2024
Y1 - 2024
N2 - Social media images are increasingly appropriated, constructed, and used as evidence. To this end, facial recognition technologies are applied to harness social media images as evidence of people’s identities, movements, and actions. This development transforms visual evidence in terms of which images are constructed as evidence and how and by whom they are constructed as evidence. It also grants access to identifying citizens on an unprecedented scale. This article develops a theoretical framework for understanding how social media images are appropriated as evidence and used as identification through facial recognition technologies. Empirically, the article studies the extensive use of social media images in identifications conducted by state actors and open-source actors, exemplified by joint efforts by the FBI and Sedition Hunters to identify the participants in the Capitol Hill siege (2021). The primary finding of the quantitative analysis is that social media images constitute almost one-third of the images used by the FBI in their identification of the rioters, while the qualitative analysis focuses on FBI’s use of leads from Sedition Hunters and this open-source actor’s key, yet obscure, application of facial recognition technologies.
AB - Social media images are increasingly appropriated, constructed, and used as evidence. To this end, facial recognition technologies are applied to harness social media images as evidence of people’s identities, movements, and actions. This development transforms visual evidence in terms of which images are constructed as evidence and how and by whom they are constructed as evidence. It also grants access to identifying citizens on an unprecedented scale. This article develops a theoretical framework for understanding how social media images are appropriated as evidence and used as identification through facial recognition technologies. Empirically, the article studies the extensive use of social media images in identifications conducted by state actors and open-source actors, exemplified by joint efforts by the FBI and Sedition Hunters to identify the participants in the Capitol Hill siege (2021). The primary finding of the quantitative analysis is that social media images constitute almost one-third of the images used by the FBI in their identification of the rioters, while the qualitative analysis focuses on FBI’s use of leads from Sedition Hunters and this open-source actor’s key, yet obscure, application of facial recognition technologies.
U2 - https://doi.org/10.1080/1369118X.2024.2358164
DO - https://doi.org/10.1080/1369118X.2024.2358164
M3 - Journal article
SP - 1
EP - 17
JO - Information, Communication & Society
JF - Information, Communication & Society
SN - 1369-118X
ER -
ID: 390511971