The Seattle-based e-commerce company's patent application filed recently describes outlines a process to allow shoppers to make a purchase by taking a photo and/or video of themselves instead of entering their account password.
The application is related to a separate patent Amazon held for a technology that allowed the authentication of a user via a photo or video, but not necessarily to complete a transaction.
The current application seeks to make it safer for shoppers to buy something online by relying on images of themselves rather than a password, which could be hard to remember and dangerous when stolen.
''The entry of these passwords … can require the user to turn away from friends or co-workers when entering a password, which can be awkward or embarrassing in many situations,'' according to the patent application.
To avoid this, some people store their password on their phone or computer so they do not need re-enter it for each purchase. However, if someone else were to get hold of that device, unpleasant consequences could follow.
''While many conventional approaches rely on password entry for user authentication, these passwords can be stolen or discovered by other persons who can impersonate the user for any variety of tasks,'' explains Amazon in its patent application.
''Further, the entry of these passwords on portable devices is not user friendly in many cases, as the small touchscreen or keyboard elements can be difficult to accurately select using a relatively large human finger, and can require the user to turn away from friends or co-workers when entering a password, which can be awkward or embarrassing in many situations.''
Further, according to Amzaon, while a number of verification systems might snap a photo of the user with facial recognition software, ''such a process provides only a certain level of additional protection.''
However, if identity-theft was a primary concern, could fraudulent payments be made with simply a photograph or video of the cardholder?
In the patent application Amazon says it could not happen with a randomly generated verification process.
''A user can be asked to perform a specific motion, gesture, or action within the field of view of the camera to help verify that the image information corresponds to a physical human user,'' explains Amazon.