Faces of the Riot facial recognition raises privacy concerns

Occurred: January 2021

Can you improve this page?
Share your insights with us

Over 6,000 facial images of people taking part in the January 6, 2021 insurrection at the US Capitol building were made available on Faces of the Riot, raising privacy concerns.

The site isolated and made public geo-tagged images of people involved in the insurrection that had been posted to Parler, the social networking service associated with Donald Trump supporters. The images had previously been downloaded by hackers after Amazon had decided to cease hosting Parler. 

Whilst each face on the site is only tagged with a string of characters associated with the Parler video in which it appeared, and faces are not named, the initiative raised privacy concerns. It was also seen to have failed to distinguish between lawbreakers and people legally attending the protests.

Databank

Operator: Anonymous/pseudonymous

Developer: Anonymous/pseudonymous

Country: USA

Sector: Politics

Purpose: Identify protestors
Technology: Facial recognition; Generative adversarial network (GAN); Neural network; Deep learning; Machine learning 

Issue: Privacy; Accuracy/reliability
Transparency: Governance; Privacy

Page info
Type: Issue
Published: January 2021
Last updated: January 2022