# AI-assisted document fraud: a column says insurers bear the burden of proof

Published 2026-11-04T11:01:38.000Z · AI Pulse · https://ai-pulse.news/policy-law/ai-document-fraud-insurers-burden-proof

*This story was written with the help of artificial intelligence, checked automatically, then reviewed by an editor before publication, under our AI policy : https://ai-pulse.news/ai-policy*

A column by Bertrand Servary of VOLTN, published on 3 November 2026, argues that generative AI makes document fraud easier to automate and that insurers bear the burden of proof, citing an Ipsos 2026 barometer in which one in five respondents had met a falsified document.

## Key points

- Bertrand Servary (VOLTN) argues in a Journal du Net column that generative AI makes document forgery accessible and increasingly automatable, without advanced technical skills.
- An Ipsos barometer for Youtrust finds that one in five working people has already encountered a falsified document, and that only 7% of companies feel fully prepared for new AI-related risks.
- According to the column, in a dispute the burden of proof rests on the insurer, and the eIDAS regulation introduces graded levels of proof.
- The column and the survey publisher name the study differently (Yousign x Ipsos versus Youtrust x Ipsos), an inconsistency the sources do not explain.

## What the column argues

In a column published on 3 November 2026 by [Journal du Net](https://www.journaldunet.com/fintech/1557315-avec-l-ia-la-fraude-documentaire-s-industrialise-les-assureurs-sont-ils-prets/), Bertrand Servary of VOLTN argues that generative AI makes document falsification accessible and increasingly automatable, without advanced technical expertise. This is the author's assessment, not a measured finding.

The column also says document fraud is no longer solely an IT matter. According to the author, it now involves compliance, legal, risk management and general management.

## Legal exposure for insurers

The column states that in a dispute the burden of proof lies with the insurer, and that the eIDAS regulation introduces a gradation of levels of proof. These are the columnist's claims. The material does not detail which provisions of eIDAS apply.

## What the survey figures say

The column cites a 2026 barometer reporting that one in five professionals has already faced a falsified or compromised digital document. The survey page words it as: «1 actif sur 5 a déjà été confronté à un document falsifié».

- Only 7% of companies consider themselves fully prepared for new AI-related risks.
- 76% of decision-makers see digital sovereignty as a strategic issue, up 11 points in one year.

The sovereignty figure is not directly tied to document fraud in the available material.

## Open questions

The barometer is conducted by Ipsos for Youtrust, publisher of a signature and verification platform that hosts the survey page. That commercial position is worth noting. The column calls the study "Yousign x Ipsos", while the publisher's page says "Youtrust x Ipsos"; the sources do not explain the difference.

The material gives no sample size, survey method or insurer-specific fraud data. The burden of proof is a claim made by the column alone, not by the survey.

## FAQ

### Who bears the burden of proof in an insurance dispute over a document?

According to the column, the insurer does, and eIDAS introduces graded levels of proof.

### How many professionals have met a falsified document?

The Ipsos 2026 barometer reports one in five.

## Sources

1. [Avec l'IA, la fraude documentaire s'industrialise : les assureurs sont-ils prêts ?](https://www.journaldunet.com/fintech/1557315-avec-l-ia-la-fraude-documentaire-s-industrialise-les-assureurs-sont-ils-prets/), Journal du Net, 2026-11-03
2. [https://yousign.com/fr-fr/guides/ipsos-yousign-2026](https://yousign.com/fr-fr/guides/ipsos-yousign-2026), yousign.com, 2026-11-04 (primary source)
