AIO Apex
en

Retailers Are Quietly Expanding Facial Recognition for Loss Prevention

Share:
Retailers Are Quietly Expanding Facial Recognition for Loss Prevention

Home Depot is facing a class action in Illinois, filed in 2025, alleging the retailer used facial recognition at self-checkout kiosks across all 76 of its Illinois stores without notice or consent — a claim that, if certified, could expose the company to statutory damages of $5,000 per violation under the state's Biometric Information Privacy Act (BIPA). Target is defending a similar BIPA suit that a federal judge in Illinois has allowed to proceed. Neither company has stopped deploying the technology while the cases work through court. That's the current state of retail facial recognition in the US: widely deployed, frequently litigated, and still mostly unregulated outside a handful of states.

The pattern matters because Illinois is doing most of the enforcement work nationally. BIPA, passed in 2008, is one of the only US privacy statutes with a private right of action — meaning individual customers, not just state attorneys general, can sue directly, and statutory damages don't require proof of actual harm. That combination has made Illinois the de facto national courtroom for retail biometrics, the same way it became the testing ground for workplace fingerprint-scanner lawsuits a decade earlier. Retailers operating nationally increasingly treat BIPA compliance as a design constraint for their entire system rather than a state-specific carve-out, because building one facial-recognition system for 49 states and a different, notice-heavy one for Illinois costs more than just building the privacy-protective version everywhere.

Not every state is waiting for a BIPA-style law to catch up. Connecticut lawmakers announced plans to introduce legislation in the 2026 session that would prohibit retailers from using facial recognition on customers altogether — a broader ban than BIPA's notice-and-consent model, closer to an outright prohibition. Grocery chain Wegmans, which uses facial recognition for self-checkout fraud prevention in select locations, issued a public statement in February 2026 defending the practice after backlash, arguing the system only flags repeat theft suspects and doesn't store biometric data on general shoppers. Whether that distinction — targeted flagging versus blanket scanning — survives contact with a state legislature remains the open legal question shaping how retailers deploy the technology going forward.

The business case retailers make is straightforward: organized retail crime losses have pushed loss-prevention budgets up across the sector, and facial recognition offers a way to flag known shoplifters at the door before they reach a shelf, rather than relying on after-the-fact loss reports. The problem is that "known shoplifter" databases are built from prior detentions and store-level incident reports, which means they inherit whatever bias existed in how those incidents were recorded and who got flagged in the first place — a concern civil liberties groups like the ACLU have raised specifically about secretive deployment at self-checkout, where customers have no way to know they're being scanned, let alone contest a false match.

Rite Aid offers the cautionary case study other retailers now study closely. The FTC banned Rite Aid from using facial recognition technology for five years after finding the company's "reckless" deployment harmed and harassed customers, including repeated false matches that led staff to confront innocent shoppers, disproportionately in stores located in majority-Black and majority-Latino neighborhoods. That enforcement action predates the current wave of retail deployments but still shapes how companies structure new rollouts — specifically, it's why most current systems emphasize narrow, opt-in matching against a small known-offender list rather than broad surveillance of every customer who walks in.

For retailers actually weighing whether to deploy this technology, three practical considerations separate the lawsuits from the lawsuit-free rollouts. First, explicit point-of-entry notice — a sign stating facial recognition is in use, not buried in a privacy policy — is the single factor that has most consistently distinguished successful BIPA defenses from ones that settled or lost; Illinois case law treats conspicuous notice plus an opt-out mechanism as substantially different from silent collection. Second, data retention windows matter more than the technology itself: a system that matches faces in real time against a known-offender list and discards the scan immediately if there's no match sits on much firmer legal ground than one that stores every shopper's facial geometry indefinitely for pattern analysis. Third, vendor contracts need explicit data-deletion and no-resale clauses — several of the BIPA suits filed since 2020 target not just the retailer but the third-party vendor, with Clearview AI named in multiple cases, because sharing biometric templates with an outside company multiplies the points of legal exposure.

For shoppers, the practical reality is that opting out individually isn't really possible once a store has deployed facial recognition at entry or checkout — there's no "turn off camera for me" setting. The leverage that exists is collective and legislative: state BIPA-style laws and outright bans like Connecticut's proposal are the only mechanisms that have actually changed retailer behavior so far, more than any individual lawsuit settlement. If your state doesn't have one, a retailer's published privacy policy is the only disclosure you're likely to get, and most don't specify facial recognition store by store. Checking for a point-of-entry sign, not a policy PDF, is currently the best real-world signal of whether a specific store is actively scanning.

Share:
Retail Facial Recognition Expands Despite BIPA Lawsuits in Illinois | AIO APEX