The collaboration between LVM Versicherung and VAARHAFT was all about AI-supported detection of falsified claims images. The aim of the collaboration was to automate fraud detection and increase process efficiency. The solution: the „Fraud Scanner“ for efficient fraud processing and prevention. The software analyses images in various formats such as JPG, PNG or PDF and detects manipulations such as generated content, image editing or duplicates.
Thanks to the API and web tool integration of the fraud scanner into LVM's internal processes, claims images can be checked as soon as they are received. Suspicious cases are automatically flagged, allowing claim handlers to be informed in a targeted manner and act without additional effort. This enables a reliable check and fraud prevention even for images without prior initial suspicion.
Important findings from the collaboration include the importance of a comprehensive process analysis prior to integration and the continuous further development of the AI models. For example, the fraud scanner will be further improved through targeted retraining during close-ups. The collaboration shows how AI technology can be used effectively to take fraud detection in the insurance industry to a new level.
Linus Kameni, VAARHAFT
„What I personally particularly appreciated was the Networking made possible by InsurLab Germany. We have been given stages, but above all we have been networked.
And we have been given the opportunity to work with many insurers in a very short space of time. to come together, to receive a lot of feedback and to actually have some pilot projects developed from this.”
Linus Kameni, CEO & Co-Founder of VAARHAFT
