It's not that hard to add points to your loyalty cards on Android Pay, but it looks like Google is mulling on an experimental feature to automate the process. 9to5google has torn the latest version of the app apart and found lines of code that hint at a feature called "Visual ID," which authenticates your loyalty points by using facial recognition. Based on the strings the publication found, you'll have to create a "face template" when you activate the feature. Participating stores that have Visual ID cameras installed will then confirm your identity when you walk in. Once the system determines that it's you, and it ascertains your location using Bluetooth, Google will send them your loyalty details.
The fact that a camera in store can take your picture and identify you sounds like a cause for concern when it comes to privacy. It's totally different from a CCTV that only captures you on cam without identifying who you are. According to the codes in the app, though, the images Visual ID cameras capture can't be accessed by the store and are deleted shortly after they're taken:
<string name="handsfree_consent_text_identity_text">"Google creates and stores a face template based on the photo you save during setup. When you're at a participating store, Visual ID will automatically confirm your identity using facial recognition technology (your face template is compared to the image captured by the in-store Visual ID camera). Images captured by the in-store Visual ID camera can't be accessed by the store and are deleted shortly after they're captured. They're not stored or saved."</string>
Another string also said that Google won't share any other details about you other than your loyalty account info. Android Pay won't use your face to authorize payments like MasterCard's new selfie security does. You can also switch off Visual ID anytime.
The feature is obviously still in its very early form, and we still don't know whether Google will ever launch it. If the big G does, it still remains to be seen how well it will work: facial recognition tech is far from perfect at this point in time, especially when it comes to recognizing non-white faces.