Skip to content


Generali France has set itself some ambitious goals in the fight against fraud and hopes for support from Artificial Intelligence and Data Science in order to optimize the detection of suspicious files.  The aim is to improve its knowledge of the types of warranty fraud that exist. Thanks to Shift Technology's Force solution, since April 2019 the two companies have been strengthening their partnership in the area of the fight against auto insurance fraud.


Having collaborated with Shift on automobile insurance since 2016, it was only natural that Generali should once again turn to this partner, which offers an insurance fraud detection solution. This time the goal is to deploy it in the property damage insurance sector.

"What Generali wants is to detect anomalies in claims files in real time, so as to analyses them and be able to process anomalies that are not really fraudulent faster. Shift's "real-time" detection module responds directly to this need", says Jérôme Burtheret, Head of Control and Reimbursement Policy at Generali France. "Actually, from the moment a claim is filed, Shift is capable of combining thousands of variables and identifying incoherences amongst the data relating to the claim and the context".

These include:

Systematization of the verification of every invoice.

Utilization of images in analyses, notably the capture of aerial images via Google Maps.

The Force solution draws its data from two sources:

Internal data: This is data delivered by the insurer to Shift.

-Claims data

-Contracts data

-Sample copy

Third party data: This is data recovered by Shift to enhance internal data.

-Government data (companies' data, demographic data...)

-Web (Google)

-Aerial imaging (Google Maps)



Shift Technology is able to cross-reference all the data gathered in order to spot any anomalies in a claim file.


Every year 125,000 claims are filed with Generali, with a ratio of around 300 fraudulent claims. In 2022 it is expected that 600 fraudulent claims will be filed. Property damage insurance for companies and private individuals corresponds to approximately 30% of P&C claims received by Generali France in terms of amount and volume. The detection and organization methods for processing of suspicious files result in 0.7% of damage claims being recognized as fraud. According to the anti-fraud agency ALFA, the most successful insurers on the market have a rate of 2%. With Shift, Generali hopes to join this group. Shift detection focuses on multi-risk home, property and business insurance, i.e. 107,000 claims out of 125,000 P&C claims. Using the Force solution, around 300 claims will be analyzed a day.

In close collaboration with Generali's technical experts, Shift's data scientists have already been mobilized to adapt the detection algorithms to the characteristics of Generali's portfolio. As from November 2019 and the end of the design phase, Force should be able to automatically generate alerts for Generali's administrators when claims are filed. "Every alert has a suspicion "score" and is substantiated by an array of indications", explains Arnaud Grapinet, Chief Data Scientist at Shift. "It is then processed by the administrators who analyze its pertinence and, where appropriate, carry out the necessary investigation or follow through the claim process after the suspicion of fraud has been ruled out. "

The automation of a part of the property damage fraud detection speeds up the processing period of claims, saving time for both insured and insurer. The IA will be able to assess the files it has erroneously detected as fraudulent and exclude them from the files to be dealt with. This is because it is estimated that automated property damage insurance fraud detection gives 89% of false positives, a high number in comparison to the percentage of recognized fraud. After the analysis has been completed, the administrator will check whether there has been any fraud and will be able to pay the policyholder more quickly and easily.

With Force, around 3,000 claims will be analyzed a day, and analysis will take place several times as each stage of the claims process moves forward.

Régis Lemarchand, a member of the Executive Committee of Generali France, who is in charge of reimbursement, says that "Shift has constantly innovated and enriched its solution over the years. Their documents and images analysis module for detecting fraud is particularly powerful. It was logical that Generali should bring Shift Technology on board its ambitious road map in the area of the fight against insurance fraud. Shift's Force solution also combats price rises in insurance policies, and at Generali we want every policyholder to pay a fair price; our aim is to ensure that the cost of these frauds is not borne by honest customers. Shift's IA will be able to provide us with support to achieve this. "

"Generali has been one of our top clients in France! We are extremely happy to extend our partnership from car insurance to damage insurance, and help Generali to attain their ambitious goals. We thank Generali for the trust they have placed in us over these years. " Jeremy Jawish, CEO and co-founder of Shift Technology


Generali France is one of the leading insurers in France today. The turnover of the company, which was established in France in 1832, reached a figure of 12.3 billion Euros in 2018. Generali France is supported by more than 9,000 workers and general agents with the aim of accompanying its clients throughout their life. It offers insurance and asset management solutions for 7.2 million people, as well as 750,000 professionals and companies.


Shift Technology delivers the only AI-native fraud detection and claims automation solutions built specifically for the global insurance industry. Our SaaS solutions identify individual and network fraud with double the accuracy of competing offerings, and provide contextual guidance to help insurers achieve faster, more accurate claim resolutions. Shift has analyzed hundreds of millions of claims to date, and was named by CB Insights to the 2018 Global AI Top 100. For more information please visit