Morrisons Deploys Everseen AI to Reduce Self-Checkout Friction

What happened

Morrisons, one of the United Kingdom's largest supermarket chains, has entered a partnership with Everseen, an AI-powered visual intelligence company, to address friction and loss-prevention challenges at self-checkout lanes. The collaboration deploys Everseen's computer-vision technology across Morrisons' self-service points of sale, with the stated aim of reducing checkout errors and improving the overall experience for shoppers using unmanned tills.

Everseen's platform uses real-time video analytics to detect scanning errors, missed items and other anomalies at self-checkout — intervening or alerting staff before a transaction completes. The partnership positions the technology not merely as a shrinkage-reduction tool but as a means of smoothing the self-checkout journey, reducing the frequency of staff interventions that frustrate customers and slow throughput.

Why it matters

Self-checkout has become one of the most behaviorally loaded touchpoints in physical retail. Customers who choose self-service typically do so to feel in control and to move quickly — yet the experience routinely undermines both goals through false alerts, "unexpected item in bagging area" interruptions and the implicit social friction of being watched for suspected theft. Every unnecessary staff intervention is a micro-moment of distrust that erodes loyalty, regardless of how politely it is handled.

By applying computer vision to distinguish genuine scanning errors from innocent behaviour, Morrisons and Everseen are effectively trying to recalibrate the trust equation at the till. From a service-design perspective, this is significant: the technology's value is not just operational (lower shrinkage, faster throughput) but experiential — reducing the number of moments where a customer feels accused, delayed or surveilled. Retailers investing in this space are implicitly acknowledging that loss-prevention and customer experience are not separate agendas; they are the same conversation.

The Renascence take

Most coverage of AI at self-checkout frames the story around theft and shrinkage. That framing misses the deeper opportunity — and the deeper risk.

The real design challenge at self-checkout is not catching bad actors; it is avoiding the treatment of good ones as suspects. Behavioral economics tells us that perceived unfairness — even a momentary, automated accusation — triggers a disproportionate emotional response that lingers long after the transaction ends. Everseen's approach is promising precisely because it shifts the system's default from suspicion to assistance. But the implementation detail will determine the outcome: if the AI reduces false positives substantially, the experience improves; if it merely moves the friction point or adds new forms of surveillance theatre, operators will have spent capital without earning trust. Customer-obsessed retailers should measure success here not by shrinkage rates alone, but by post-checkout satisfaction scores and the rate of unprompted staff interventions per transaction.