Scandit Launches Vision AI to Curb Self-Checkout Theft

The software, dubbed Self-Checkout Loss Prevention, uses existing store camera infrastructure to identify suspicious patterns, such as skipped items or goods left in a trolley. When an anomaly occurs, the system triggers a soft on-screen nudge, allowing the shopper to correct the mistake independently. According to the company, this automated approach recovers or deters more than 75% of losses associated with self-service kiosks.

Christian Floerkemeier, CTO and co-founder of Scandit, stated that the goal is to remove the trade-off between checkout speed and security. By processing data directly at the station, the system avoids the latency and high costs typical of legacy hardware-heavy setups. Furthermore, the architecture is designed for privacy, as it analyzes transaction activity rather than biometric markers, ensuring compliance with GDPR and the EU AI Act.

This launch follows the company's July 2026 rollout of an Age-Verified Self-Checkout tool. By combining both systems, retailers can automate the two most frequent sources of friction that require associate intervention. Greg Buzek, president of IHL Group, noted that repurposing existing cameras to resolve honest errors while deterring intentional theft provides a significant operational advantage for retailers struggling with inventory shrinkage.