Walmart says it is now using AI-powered cameras in more than 1,000 stores to prevent theft at checkouts, in a program internally called Missed Scan Detection
Hayley Peterson / Business Insider :
Context & Ripple Effects
This lands mid-arc in Walmart's store-automation push: after testing shelf-scanning robots in 50+ stores in 2017 and expanding floor-scrubbing bots and Pickup Towers across the US in April 2019, the company is now pointing cameras at its own checkouts. Missed Scan Detection turns loss prevention from a human-watching-humans job into a software problem at scale.
The stakes go beyond shrinkage. A year later, staff would claim the same class of AI anti-shoplifting tech misidentifies innocuous behavior while missing actual theft, making this rollout the deployment whose accuracy disputes define how retail enforcement AI gets judged.
First-order effects
- Checkout associates and loss-prevention staff in 1,000+ stores shift from spotting theft themselves to responding to camera-flagged missed scans, changing what their role looks like on the floor.
- Shoppers at self-checkout face a new surveillance layer where every unscanned item can trigger an intervention, with Walmart's deterrence case resting on the system's hit rate.
Second-order effects
- Rivals like Target, which competes with Walmart on both price and checkout experience, face pressure to match AI-based loss prevention or absorb relatively higher shrink costs.
- If false-positive rates prove high, as staff later alleged, Walmart absorbs the cost in employee time spent resolving wrongful accusations and in customer friction at exactly the moment it is investing in store experience.
Third-order effects
- Retail enforcement is becoming an AI accountability battleground: accuracy disputes over these cameras foreshadow the governance scrutiny that follows any algorithm deciding who gets accused of theft.
- Walmart's own later framing — that AI changes how people work rather than how many — suggests this program is a template for reassigning frontline labor around machine-flagged exceptions rather than cutting headcount outright.
The trend: Big-box retail is converting physical-store oversight — shelves, floors, and now checkouts — into AI-monitored surfaces, with accuracy disputes shaping how far enforcement applications can spread.