Looking at a vulnerability, with the potential of getting an attacker more videos than they can investigate, the researchers decided to take it one step further by using a computer vision technology that is capable of video analysis. We could have used Google Vision, Azure Computer Vision, or any other service, but we decided to go with the excellent
Amazon Rekognition service. Rekognition can be used to automate the analysis of these recordings and extract information that could be useful for malicious actors.Rekognition can scan an unlimited number of videos and detect objects, text, faces, and public figures, among other things.
To further demonstrate the impact of this vulnerability, the researchers showed how this service could be used to read sensitive information from computer screens and documents visible to the Ring cameras and to track people’s movements in and out of a room.
