Amazon has started rolling out Familiar Faces for Ring video doorbells, a facial recognition feature that tags frequent visitors and sends personalized alerts, and the move has split opinion between convenience-minded users and privacy advocates who see new risks. This article lays out how Familiar Faces works, the data it stores, past Ring security concerns, legal pushback, and other Ring AI features like Video Descriptions and Greetings that take a different approach. Read on for the nuts and bolts, the criticisms, and practical cautions to consider if you use Ring devices.
Familiar Faces uses AI to identify people who commonly appear at your door and replaces generic notifications with labeled alerts such as a family member’s name. Users can create a library of up to 50 faces and tag them from camera event clips, which then lets the system recognize those people on future visits. That shifts a simple motion alert into a personalized notice, which many people will find convenient.
The feature requires manual activation in the Ring app and gives owners tools to name, merge, edit, or delete faces from the Familiar Faces library. Amazon says unnamed faces are removed automatically after 30 days, but once a face is labeled the associated data stays until the owner deletes it. That persistence means the owner controls deletion, while the company retains processed information unless users act.
Privacy groups have been blunt about the risks, and one critic put it this way: “When you step in front of one of these cameras, your faceprint is taken and stored on Amazon’s servers, whether you consent or not. Today’s feature to recognize your friend at your front door can easily be repurposed tomorrow for mass surveillance. It is important for state regulators to investigate.” Those concerns connect facial identification to broader debates about biometric data and how it can be reused.
Ring’s record amplifies those worries. The company has a history of partnerships and integrations with law enforcement tools, and it has faced security issues, including a 2023 FTC fine of $5.8 million after investigations found employees and contractors had wide access to customer videos. Past problems like exposed home locations via the Neighbors app and leaked account credentials feed the argument that adding facial recognition increases potential harm rather than reducing it.
Regulatory pressure is already shaping where Familiar Faces can be offered, with certain jurisdictions blocking the feature under stricter biometric privacy laws. That patchwork of rules reflects an uneasy balance between consumer convenience and statutory protections, and some lawmakers have urged Amazon to stop the rollout entirely. The limits show regulators treating biometric ID technology differently from routine smart-home features.
Amazon insists biometric processing happens in the cloud and that the company does not use the stored face data to train its core AI models, and it says it cannot always map where a face appears across all locations even if asked. Still, critics point to features like Search Party, which scans neighborhoods, as evidence that similar tools can be repurposed for broader searches. Those similarities make people wary of future expansions.
Ring’s other AI offerings take a less identity-focused tack. Video Descriptions is a generative AI feature that summarizes detected motion into plain language, producing alerts such as “A person is walking up the steps with a black dog” and “Two people are peering into a white car in the driveway.” This kind of description highlights activity rather than naming people, and it can help owners triage alerts without storing identity markers.
A different AI capability, Greetings inside Alexa+, gives doorbells a conversational layer that interacts with visitors based on visual context and user rules. Greetings can direct delivery drivers where to leave packages, ask about signatures, or tell solicitors you are not available, and those interactions are saved so owners can review messages later. Because it relies on video descriptions of actions and objects rather than facial identity, it is framed as a privacy-safer option, though it is not immune to mistakes when context is ambiguous.
If you use Ring devices, exercise caution before turning on Familiar Faces: avoid using full names, regularly delete faces you no longer need, and weigh whether the convenience of labeled alerts is worth the record of who visits your home and when. For many users, checking live video manually or relying on activity-based descriptions may provide adequate security without building a stored catalog of faces. Smart home tools offer choices, and choosing privacy-minded settings remains a practical approach for those concerned about biometric tracking.
Deciding to enable Familiar Faces is a personal trade-off between fewer alerts and more detailed visitor records, and homeowners should consider local laws, Ring’s history, and the alternative AI features available before changing their settings. If you keep the feature off, Video Descriptions and Greetings can still add value without associating names with faces, giving a middle path for users who want smarter alerts but not stored identities.
