Richard Weurding, general director of the Association of Insurers: "Those who commit fraud disadvantage everyone who pays premiums fairly. That's why insurers don't just look at suspicious claims but also at abuse when taking out insurance. Early action to prevent fraud is at least as important as detecting fraud afterwards."
In 2024, insurers identified 9,070 fraud cases and saved 95.6 million euros. In 2025, the number of detected frauds thus increased, while the savings remained virtually unchanged. This concerns established fraud in both insurance applications and claims against non-life and life insurers. Fraud with health insurance is excluded from these figures. The figures do not show how much fraud is committed in total or goes unnoticed.
Control starts with the application
Of all frauds detected in 2025, 69 percent were discovered when applying for insurance. By recognising deception at that moment, insurers can prevent insurance from being established that later, for example, leads to unfounded claims. The approach remains important for business insurance as well: in 2025, more than 11 percent more fraud cases were detected there than in 2024.
A stain that exists only in the photo
For example, there is a case where an insured claimed damage to a carpet. The accompanying photo showed a stain that had been added with AI. The insurer rejected the claim and registered the person involved in the joint warning system. In another case, an insured person deliberately increased the claim by over 1,000 euros after the theft of a motorcycle. After investigation, the insured admitted to having forged documents. That claim was also rejected and the policy was terminated.
The examples show that AI is used in insurance fraud. AI also simultaneously offers insurers opportunities to recognise deviations and suspicious patterns (fraud detection). However, a signal is not yet proof of fraud: careful investigation for gathering evidence and human evaluation remain necessary.
Sharing knowledge to tackle digital fraud
Through i-CERT, the cyber crisis team in which insurers and the CBV collaborate, the sector shared 302 recommendations on cyber incidents and threats in 2025, compared to 258 in 2024. For example, after the discovery of misuse of an old customer service number for fake payment requests, a warning could quickly be spread to other insurers. The CBV also collaborated with its partner Fox-IT to raise awareness and provide protection advice in response to an international group of cybercriminals targeting insurers.
Better cooperation with police and justice
The Association of Insurers supports insurers with knowledge, information, and advice and serves as a link to law enforcement agencies. In 2025, the centre handled 140 information requests from police and justice. Earlier, the Public Prosecution Service, the police, and the Association jointly investigated how insurers' investigations under strict conditions could contribute to criminal investigations. Weurding: "Fraudsters look beyond just one insurer. Therefore, we must continue to share knowledge and signals and optimise cooperation with police and justice. With clear agreements on information exchange and linking fraud investigations to criminal investigations, we can act more effectively together. Care, privacy, and the rights of those involved must always be safeguarded."
Measures and due diligence
In case of proven fraud, insurers take measures. Depending on the case, they can, among other things, reject a claim, terminate the policy, reclaim amounts paid out or research costs, and file a report. In 2025, 3,588 parties were included in the External Referral Register (EVR), the joint warning system. This does not happen automatically in every fraud case: strict conditions and individual balancing of interests apply to registration. A registration can take a maximum of eight years.
Privacy rules also apply when sharing information about fraudsters. Since April 1, 2026, a renewed Protocol for the Financial Institutions Incident Warning System (PIFI) has been in effect. The Dutch Data Protection Authority has granted new licenses for eight years, allowing financial institutions to continue sharing information about incidents and fraud instigators within the applicable conditions.
Read our factsheet
Here you can read the factsheet (in Dutch) containing all the figures and examples of fraud.
Spokesperson for Damage
For questions, please contact:
Veilig wonen & ondernemen, cyber, letsel, verkeer, klimaat/duurzaamheid, reizen, aansprakelijkheid, rechtsbijstand, volmacht, verzekeringsfraude- en criminaliteit en zakelijke markt.