Privacy

On Facial Recognition and Identity Proofing

Wired has a good piece on the IRS in the US caving to public outcry and ditching its integration with ID.me - a service that was supposed to verify identities (by matching video selfies to existing records). It’s understandable why this would cause concerns given that facial recognition is rife with false matches, biases and a reputation for invasiveness. With fraud being a pressing issue now when a majority of us (at least in Australia) access nearly every civic service online, governments are going to want to think about how they balance policy, privacy and messaging.

Unfortunately, the landscape at the moment is messy and populated by a number of third-party vendors still finding their feet in an area where privacy and policy concerns are outweighed by sexier usability and convenience use cases.

“The fact we don’t have good digital identity systems can’t become a rationale for rushing to create systems with Kafkaesque fairness and equity problems.” - Jay Stanley, ACLU

It’ll be interesting to see how Australia’s Trusted Digital Identity Framework (TDIF) will look to address some of these inherent problems through a continuous expansion of its standards.