Insurance fraud is becoming increasingly sophisticated, with fraudsters exploiting small but repeatable tactics that are difficult for investigators to catch. One such tactic is the reuse of photos across different claims — a method that, without automation, can easily slip through manual review processes and result in significant losses for insurers.
Shift’s AI-based document fraud detection leveraged image similarity scoring to detect that the same refrigerator photo appeared in multiple claim submissions. The system compares image features to generate a similarity score that flags potential duplicates, even if images have been cropped, resized, or slightly altered. Once the reused photo was detected, the AI automatically generated an alert. Further cross-analysis of the claims uncovered another connection: the same intermediary was involved in each of the fraudulent cases. By combining computer vision-based photo analysis with network link detection, Shift’s AI provided insurers with clear, actionable evidence to block the claim before payment and dismantled the broker’s scheme.
This video is part of a series of interesting cases presented by our insurance-focused team of data scientists. For more examples of Shift Technology's AI-driven results, browse the AI in Action library