Document automation (IDP) requires authentication

Insurance companies and banks, as well as the public sector, digitise hundreds of thousands of documents every day, which form the basis for decisions such as payments. Unfortunately, fraud prevention, i.e. distinguishing between genuine and forged documents, is still not automatically taken into account in digitisation projects. In the hope of increasing efficiency, companies are opening up a huge gateway for fraud.

In order not to reduce the efficiency gains of digitisation through fraudulent activities and to pass on the costs to honest customers, the authenticity check of documents should always be taken into account in document automation (IDP). With ICO.Fraud, ICO-LUX has an innovative product for fraud defence and document authentication. With ICO.Fraud, forged documents are recognised fully automatically and in real time - thanks to smart vision technology and cognitive automation.

How ICO.Fraud redefines fraud defence

Fraud prevention with ICO.Fraud is based on high-performance image recognition. The software recognises fully automatically what type of document it is, who the document creator is and to whom it is addressed. The technology behind it, ICO.Match, is part of ICO.Fraud and can also be used separately for information extraction. Put simply, the information contained in a scanned or photographed document is converted into machine-readable, structured data records in a multi-step process: Perspectives are straightened, background noise is removed, table structures and text as well as spaces are recognised. Everything is assigned to recurring patterns of documents that were written by the same person.

Next-level fraud detection: Highly performing and precise

High-calibre data extraction with the ICO.Match technology, which is always integrated in ICO.Fraud, recognises and contextualises all data on documents. This structured data is then further processed in ICO.Fraud: With the help of a sophisticated set of rules at the heart of the solution, even the smallest layout deviations can be detected, analysed and evaluated in comparison to other documents of the same origin. Additional features such as a patient calendar, which compares treatment data with practice opening times and determines distances between any treatment locations, as well as analysing the consecutive invoice numbers of the same practice from other documents are also included in the fraud check.

ICO.Fraud scores with a pleasant user interface. In line with the European Union's Explainable AI Act, the software not only flags suspicious documents, but also always makes it transparent why something was recognised as suspicious, so that the final decisions always remain with the employees in the fraud departments.