Bill W:
Chad G:
#3 of 15 in a series: Driver Criticism, Voice Recognition, Privacy Rights and Deactivation
Story Idea #013 – Clearview AI: Why Publicly Posted Images Weren’t Fair Game for Biometric Identification in Canada
If a photograph is publicly visible on the Internet, does that mean anyone may turn it into a biometric identifier and use it to discover who you are? Canadian privacy regulators said no in their investigation of Clearview AI.
Clearview’s position included the argument that the images it collected were already publicly accessible. Regulators rejected the idea that public visibility automatically permits unrestricted biometric reuse, finding serious problems involving consent and appropriate purpose.
We want to explain why that distinction matters far beyond facial recognition. Posting something publicly can expose it to viewers without necessarily granting permission for every possible technological transformation of it. In our wider gig-driver research, that principle becomes relevant to a driver who posts public audio or video while trying to remain pseudonymous.
The Clearview case is not identical to a hypothetical voice-matching system. We will not pretend it is. But paired with Canada’s Rogers voiceprint decision, it provides a useful framework for asking where ordinary public content ends and biometric identification begins.
Curious about the raw brainstorming behind this article?
