Ring’s Familiar Faces feature triggers privacy concerns

Amazon’s new Familiar Faces feature for Ring cameras is intended to make notifications more intelligent. Instead, it is provoking a wave of discomfort from owners and even people just walking past front doors. Now, as the feature comes closer to being released, that same age-old compromise — of convenience versus consent — has landed back on our doorsteps.

How Ring’s Familiar Faces feature actually works

Familiar Faces allows Ring users to tag and name people the camera sees — family members, friends, regular visitors — so that in future alerts you can see who’s likely at the door and search recordings by face. The idea: quicker context and less generic “motion detected” pings.

Amazon says the feature will begin rolling out starting in December on new Ring doorbells and cameras. It is off by default and you will need the Ring Home Premium plan. The company has not specified publicly how long it keeps facial data or where templates are stored, either locally or in the cloud, questions privacy advocates believe are important.

Why users and bystanders are worried about consent

The immediate concern is consent. Customers are expected to comply with any state or local rules that mandate obtaining consent before identifying someone using facial recognition, a spokesperson told The Washington Post. That puts legal risk on consumers who may not know the difference between recording a porch and creating a searchable face database of everyone who comes near it.

Since many doorbells record public sidewalks and shared hallways, passers-by — delivery workers, neighbors, children on their way to school — can be swept up into face-tagging without signing any release forms. The Electronic Frontier Foundation and others have long warned that private facial recognition can normalize pervasive identification in everyday life, sometimes without clear recourse for people who don’t want to play along.

There’s also the accuracy problem. The National Institute of Standards and Technology has recorded demographic performance differentials for numerous facial recognition algorithms. Even relatively low error rates can lead to pervasive mislabeling when cameras are running all day, which increases the risk of false accusations or embarrassing neighbor feuds.

Legal patchwork and where Ring won’t launch the feature

Amazon says that Familiar Faces will not be available in Illinois, Texas and Portland, Ore., where robust biometric privacy rules restrict the collection and storage of face data without clear permission. Illinois’ Biometric Information Privacy Act has become a model statute that allows residents to sue over misuse of facial templates; the strongest punishment under BIPA remains Facebook’s $650 million settlement.

Elsewhere, rules can span from broad notice-and-consent laws to near silence. That jumble of patchwork rules adds a layer of uncertainty for homeowners who want to play by the rules, and for bystanders who might have fewer protections depending on where they are in their ZIP code.

Data stewardship questions and Ring’s track record

Trust depends on how face data is treated. Will model templates be end-to-end encrypted? Are users able to audit or delete them easily? Will they enable training of algorithms across households? Not all of these questions have been publicly answered by Amazon, adding to the skepticism.

History doesn’t help. The Federal Trade Commission announced a $5.8 million settlement with Ring in 2023 on claims that employees and contractors had illicitly viewed videos of customers and that security lapses opened the door for customers to be hacked. The order also imposed major new privacy and security requirements. On a separate front, while the company discontinued a program that allowed police departments to request footage through its neighborhood app, law enforcement agencies can still seek video recordings of crimes via legal process.

Scale magnifies risk. Market researchers including Parks Associates estimate that video doorbells now adorn many millions of United States homes, with Ring being one of the perennial market leaders. A seemingly minor feature in one home can have citywide implications when multiplied across neighborhoods.

How it compares to competitors from Google and Apple

Ring isn’t the first to think of this kind of thing. Google’s Nest cameras have a similar feature for detecting faces, and more recent Nest models analyze and store your facial scans on-device locally as well as encrypted backups. Apple’s HomeKit Secure Video uses on-device intelligence that is tethered to the Photos library, so as much facial recognition as possible can happen on a home’s hardware.

The technical details are important here; local processing, minimal data retention and clear rules for deletion all limit exposure if an account is compromised or a cloud system fails. Without concrete promises on those fronts, we’re going to resist any form of face tagging.

What consumers can do now to protect privacy with Ring

If you’re a Ring owner, maybe think about leaving Familiar Faces off when it’s launched and/or reading the fine print.

• If you do use it, ensure that regular visitors give written permission before filming them at your door.
• Point cameras away from public areas to avoid recording people passing by.
• Set up privacy zones and limit how long the system keeps a library of facial data.
• Enable two-factor authentication, and consider using Ring’s optional video end-to-end encryption (it can disable some advanced features).

For everyone else, now’s the time to ask neighbors and building managers what their camera setup is like. As cities squabble over guardrails and companies vie for smarter alerts, the difference between useful context and involuntary surveillance will come down to specific design choices — where data is stored, who has access to it and how easy it is to delete.