Document Fraud in France 2026: Data, Trends and Risk Index by Sector

Document fraud costs French businesses between 4 and 6 billion EUR per year once undetected fraud is included — three to four times the official figure of 1.6 billion EUR in declared losses. That gap between reported and actual losses is the most expensive blind spot in French compliance. The ACFE estimates that 63% of fraud incidents are never detected, and the median time to discovery still exceeds 14 months in organisations without automated controls.

This raises a structural question: how do you assess a risk when most of it escapes measurement? To answer it, CheckFile built the CheckFile Document Risk Index — a proprietary framework that cross-references sector, document type and risk factors to produce an actionable vulnerability score. This article presents the full state of document fraud in France in 2026, the detail of this methodology and an in-depth sector-by-sector analysis.

For general document fraud statistics and trends, see our dedicated report. For a comprehensive overview of fraud data sources and methodologies, consult our fraud data guide. This article focuses on the French market: structural analysis by sector and the scoring framework.

Why France matters for compliance professionals

France is the eurozone's second-largest economy and a consistent regulatory pacesetter within the EU. French compliance frameworks — from the Sapin II anti-corruption law to the transposition of AMLD6 — frequently set the template for EU-wide harmonisation. For compliance teams operating across Europe, understanding the French fraud landscape is not optional: it is a leading indicator of where regulation and risk are heading across the bloc.

France also presents unique fraud patterns. The country's competitive rental market generates a volume of forged income and address documents that is unmatched in other EU jurisdictions. Its extensive subcontracting chains in construction and public procurement create specific corporate document fraud vectors. These France-specific dynamics make a dedicated analysis essential.

2026 overview: document fraud in France by the numbers

Aggregated data

The document fraud landscape in France in 2026 is drawn from five primary sources: the annual report from TRACFIN (France's financial intelligence unit, equivalent to the UK's NCA Financial Intelligence), control data from the DGCCRF (the consumer protection and anti-fraud authority, comparable to Trading Standards), the Payment Security Observatory report from the Banque de France, the ACFE Report to the Nations 2024 and the Europol SOCTA 2025 report on organised crime.

The key indicators converge:

• 1.6 billion EUR in annual declared losses by French businesses (Banque de France, 2026 estimate).
• 4 to 6 billion EUR in estimated real losses including undetected fraud (ACFE extrapolation, 63% non-detection ratio).
• 71% of businesses targeted by at least one document fraud attempt in 2025 (PwC Global Economic Crime Survey).
• 38% of detected forgeries show AI-generation markers, up from less than 2% in 2021 (DGCCRF, 2025 control data).
• 36,500 TRACFIN reports linked to document fraud in 2026, a 10% year-on-year increase.
• 72-day average detection time, down steadily from 112 days in 2021 as automated solutions gain adoption.

France in the European context

France ranks among the top European markets for document fraud by volume, behind Germany and level with Italy. The Europol SOCTA 2025 report positions France as the third-largest European market for forged documents, with a distinctive feature: a high share of fraud involving income justification documents (payslips, tax notices), directly linked to the requirements of the French rental market.

Sources: Europol SOCTA 2025, ACFE Europe, respective national estimates.

The gap with Germany reflects the size of the German financial market and automotive leasing sector. The higher deepfake share in the Netherlands reflects the advanced digitisation of Dutch administrative processes, which creates a favourable environment for digital forgeries.

Evolution 2020-2026: year-by-year data

Sources: Banque de France, TRACFIN annual reports, ACFE Report to the Nations 2024, PwC Global Economic Crime Survey.

Three major inflection points emerge from this timeline:

1. 2020-2021: the COVID shock. The forced dematerialisation of document exchanges tore open a massive breach. Physical controls — in-branch verification, visual inspection — were eliminated overnight. Losses jumped 22% in a single year.

2. 2022-2023: the generative AI democratisation. The availability of models capable of generating realistic images, text and layouts reduced the production cost of a forged document by 90%, according to the Payment Security Observatory. The deepfake share of detected forgeries leapt from 5% to 12% in one year.

3. 2024-2026: industrialisation. Europol now identifies 14 organised networks specialising in forged document production targeting France. These networks offer complete "packages" (identity + address + income) for 500 to 3,000 EUR. Fraud has shifted from cottage industry to industrial scale.

CheckFile Document Risk Index: proprietary framework

Why a sector-specific risk index?

Aggregate statistics mask very different realities across sectors. A bank and a property management company do not face the same types of forgeries, the same volumes, or the same financial impact. The CheckFile Document Risk Index was designed to provide a granular measure of document fraud risk by sector and by document type.

Methodology

The index is built on three cross-referenced analytical axes:

Axis 1: Sector Six sectors analysed, selected for their regulatory exposure and document processing volume: Banking, Real Estate, Insurance, Construction/Subcontracting, Leasing/Finance, Public Sector.

Axis 2: Document type Five document families most frequently forged in France: Proof of address, Payslip, Identity document, Kbis/Trade register extract (the French equivalent of a Companies House certificate), Financial statements.

Axis 3: Weighted risk factors Each cell (sector x document) receives a score from 1 to 10, calculated from three weighted factors:

The overall sector score is the weighted average of scores by document type, adjusted by the relative volume of each document type within the sector in question. A score of 10 indicates maximum risk; a score of 1 indicates minimal risk.

Risk matrix: sector x document type

Reading the matrix

The two highest-risk sectors are Banking (7.6) and Leasing/Finance (7.4). In both cases, risk is distributed uniformly across all document types, meaning these sectors are exposed on every front simultaneously.

Real Estate (6.2) presents a highly concentrated risk profile: proof of address and payslips reach the maximum score of 9, but risk on Kbis and financial statements is low. This is a sector where fraud is massive but predictable in its forms.

Construction/Subcontracting (5.4) shows a moderate overall score, but with a marked peak on Kbis/trade register documents (8) and financial statements (7). Fraud in this sector primarily targets the identity and solvency of subcontractors — a specific risk vector documented in our construction subcontractor compliance analysis.

The Public Sector (5.8) stands out with a high score on Kbis/trade register documents (9), reflecting the volume of fraud in public procurement. Forged Kbis extracts allow participation in tenders using fictitious or deregistered companies.

Sector-by-sector analysis

Banking — Overall score: 7.6

Risk profile: Maximum and homogeneous exposure. Banking is the sector where risk is highest and most diversified. Forged payslips (score 9) are the primary vector, used to obtain consumer credit and mortgage loans. Falsified financial statements (score 8) target business lending.

Most common fraud types:

• Forged payslips with inflated income for credit applications
• Falsified proof of address for account openings
• Manipulated financial statements for business financing
• Identity theft for fraudulent wire transfers

Detection challenge: The volume of applications processed (several thousand per branch per month) makes systematic manual review impossible. The ACPR (France's banking and insurance supervisor, equivalent to the PRA/FCA) sanctioned 12 institutions in 2025 for failures in document controls, with cumulative fines of 18.4 million EUR. The AMLD6 directive further reinforces vigilance obligations.

Trend: Rising. The accessibility of generative AI tools increases the volume of attempts. Per-incident losses remain stable, but incident numbers grow 12-15% per year.

Real Estate — Overall score: 6.2

Risk profile: Highly concentrated on two document types. Real estate is the sector where forged proof of address and payslips are most frequent. One in five rental applications submitted in the Ile-de-France region (Greater Paris) contains at least one manipulated document, according to data from major property management firms.

Most common fraud types:

• Payslips with inflated income (typically 20-40% above actual)
• Forged tax notices (avis d'imposition — the French equivalent of a P60/tax return)
• Fabricated proof of address (fake rent receipts)
• Forged permanent employment contracts (CDI — the standard French open-ended contract)

Detection challenge: Pressure in the rental market (particularly in Ile-de-France, Lyon and Bordeaux) pushes applicants to embellish their files. Estate agents and property managers, under commercial pressure, rarely have access to automated verification tools. The detection rate in this sector is estimated at 28% — the lowest of all sectors analysed.

Trend: Stable in volume, but rising in sophistication. Crude forgeries (Photoshop retouching) are progressively being replaced by documents entirely generated by AI.

Insurance — Overall score: 6.0

Risk profile: Spread between identity documents and financial documents. Insurance is exposed to two distinct vectors: identity theft at subscription (identity document, score 8) and financial document manipulation in claims processing (financial statements, score 7).

Most common fraud types:

• Identity theft for policy subscription
• Forged valuation certificates to inflate compensation
• Falsified financial statements for professional liability insurance
• Forged proof of address to alter premium calculations

Detection challenge: Insurance fraud often surfaces late, at the point of a claim. The time between fraudulent subscription and detection can reach 18 to 24 months. AI-based detection techniques can reduce this delay by screening documents at the subscription stage.

Trend: Moderately rising. Synthetic identity fraud — creating entirely fictitious profiles from deepfake documents — is the primary growth vector.

Construction / Subcontracting — Overall score: 5.4

Risk profile: Concentrated on corporate documents. Construction is the sector where forged Kbis documents (score 8) and falsified financial statements (score 7) are most frequent. Fraud primarily targets the qualification of insolvent or fictitious subcontractors.

Most common fraud types:

• Forged Kbis to conceal company deregistration or insolvency proceedings
• Falsified URSSAF certificates (French social security compliance attestations, similar to HMRC tax clearance certificates) and compliance certificates
• Manipulated financial statements to meet solvency criteria
• Forged professional qualification certificates

Detection challenge: Cascading subcontracting chains (up to four or five tiers) make exhaustive verification difficult. Main contractors verify first-tier subcontractors but rarely check lower tiers. Supplier vigilance attestations are a first line of defence, but their verification remains largely manual.

Trend: Stable. Fraud volume is constant, but strengthened vigilance obligations (Sapin II anti-corruption law, duty of vigilance legislation) are pushing main contractors toward automated verification solutions.

Leasing / Finance — Overall score: 7.4

Risk profile: Second-highest risk sector, with particularly strong exposure on financial statements (score 9) and payslips (score 8). Leasing combines the risks of bank lending and professional financing.

Most common fraud types:

• Falsified financial statements to obtain business financing
• Forged payslips for personal vehicle leasing
• Manipulated Kbis to conceal insolvency proceedings
• Forged proof of address for finance lease agreements

Detection challenge: Leasing operates under intense commercial pressure, with signature targets that incentivise accelerated checks. The average cost of a fraud incident in leasing reaches 86,000 EUR for falsified financial statements — the highest per-incident amount of all sectors analysed. Compliance requirements are tightening under the combined effect of AMLD6 and DORA.

Trend: Strongly rising. B2B financing concentrates 42% of total losses linked to document fraud. Organised networks increasingly target this sector due to the high values at stake.

Public Sector — Overall score: 5.8

Risk profile: Dominated by Kbis/trade register fraud (score 9) in the context of public procurement. Identity documents (score 7) are the second vector, linked to social benefits fraud.

Most common fraud types:

• Forged Kbis to participate in public tenders with fictitious companies
• Forged non-exclusion certificates
• Identity theft for social benefits
• Falsified tax and social security attestations

Detection challenge: Government agencies process enormous volumes of documents with limited human resources. Kbis verification is often point-in-time (at the moment of tender submission) rather than continuous. The French Interior Ministry's programming law (2025) increased resources for the OCLCIFF (the specialised fraud investigation unit) by 25%, but the impact on detection will not materialise until 2027-2028.

Trend: Stable in volume, with potential for decline in the medium term as the European Digital Identity Wallet (eIDAS 2.0) and secure digitisation of public procurement are deployed.

2020-2026 evolution and 2027-2028 projections

Structural metrics

Beyond the annual figures presented in the overview, three structural metrics define the trajectory of document fraud in France.

The ratio of detected to actual fraud is deteriorating. Although the detection rate is improving (from 24% in 2020 to 41% in 2026), the total volume of attempts is growing faster. In absolute terms, undetected fraud continues to increase.

The cost of producing a forgery has collapsed. According to the Payment Security Observatory, the average production cost of a forged document has fallen from 150-300 EUR in 2020 (Photoshop, graphic skills required) to 10-30 EUR in 2026 (AI tools, automated templates). This 90% reduction in the cost of entry dramatically widens the pool of potential fraudsters.

Detection time is shrinking, but remains high. The drop from 112 days in 2021 to 72 days in 2026 represents significant progress, but this delay remains incompatible with the real-time compliance requirements imposed by AMLD6 and DORA.

2027-2028 projections

Projection assumptions: 7-8% annual loss growth rate (a slowdown from the 10% of 2023-2025), accelerating adoption of AI detection solutions (detection rate rising from 41% to 52% in two years), continued digitisation trend.

Two factors could significantly alter these projections:

• Downside factor: Effective deployment of the European Digital Identity Wallet (eIDAS 2.0) could dramatically reduce identity and proof-of-address fraud from 2028, provided the adoption rate exceeds 40% of the population.
• Upside factor: Continued improvement in generative AI models could render deepfake documents indistinguishable from originals, even for current detection systems, necessitating a technological leap in verification methods.

Methodology and sources

Data sources

The analysis presented in this article draws on the following sources:

• ACFE Report to the Nations 2024: global survey on occupational fraud, 1,921 cases analysed across 138 countries. Data on non-detection rates, discovery timelines and costs by sector.
• TRACFIN annual activity report 2025: official French data on suspicious activity reports, fraud typologies and affected sectors. TRACFIN (Traitement du Renseignement et Action contre les Circuits Financiers clandestins) is France's financial intelligence unit, responsible for receiving, analysing and forwarding suspicious transaction reports. It operates under the Ministry of Economy and Finance.
• Banque de France — Payment Security Observatory: statistics on payment fraud and associated document fraud.
• DGCCRF control reports 2025: data from field inspections on fraudulent practices. The DGCCRF (Direction Generale de la Concurrence, de la Consommation et de la Repression des Fraudes) is France's consumer protection and anti-fraud enforcement directorate.
• Europol SOCTA 2025: mapping of organised crime across Europe, including forged document networks.
• PwC Global Economic Crime Survey (2022-2025): survey of 5,000 businesses across 99 countries on the incidence of economic crime.
• Euler Hermes — Business Fraud Study: data on average cost per incident by sector.

CheckFile Document Risk Index calculation

The Index is calculated from aggregated and anonymised data drawn from three sources:

1. CheckFile operational data (2024-2026): document volumes analysed, detection rates by document type and sector, fraud typologies identified. This data is collected in anonymised form in compliance with GDPR.
2. Public data: TRACFIN, DGCCRF, ACFE and Europol reports cited above.
3. Qualitative client feedback: structured interviews with compliance officers from 35 companies across the six sectors analysed.

The score for each cell (sector x document) is calculated using the formula:

Score = (Frequency x 0.40) + (Financial impact x 0.35) + (Detection difficulty x 0.25)

Each factor is normalised on a scale of 1 to 10 from raw data. The overall sector score is the average of scores by document type, weighted by the relative share of each document type in the sector's total fraud volume.

Limitations and caveats

• Undetected fraud data is by definition estimated. The 63% non-detection ratio (ACFE) is a global average that may vary by sector and country.
• CheckFile operational data reflects the profile of our client base, which over-represents the banking and leasing sectors. Scores for construction and the public sector rely more heavily on public data.
• The 2027-2028 projections are linear extrapolations adjusted by qualitative factors. They do not constitute forecasts.
• This framework is updated annually based on the most recent aggregated anonymised data.

For a comprehensive overview, see our document fraud data trends guide.

FAQ

What is the real cost of document fraud in France in 2026?

The official cost, based on declared and detected losses, is estimated at 1.6 billion EUR per year (Banque de France, TRACFIN). However, the ACFE estimates that 63% of fraud incidents are never detected. Including this invisible fraud, the real cost is estimated at between 4 and 6 billion EUR per year for French businesses. This gap is explained by the absence of automated controls in the majority of organisations, a still-high average detection time (72 days) and the difficulty of quantifying indirect losses (reputational damage, remediation costs).

What is the CheckFile Document Risk Index?

The CheckFile Document Risk Index is a proprietary framework that assesses document fraud risk by sector and by document type on a scale of 1 to 10. It cross-references three weighted factors: fraud attempt frequency (40%), average financial impact per incident (35%) and detection difficulty (25%). The score is calculated from CheckFile operational data, public reports (ACFE, TRACFIN, DGCCRF, Europol) and qualitative feedback from 35 client companies. In 2026, the highest-risk sectors are banking (7.6/10) and leasing (7.4/10).

Which sectors are most exposed to document fraud in France?

According to the CheckFile Document Risk Index, the most exposed sectors are banking (score 7.6/10) and leasing/finance (7.4/10), followed by real estate (6.2/10) and insurance (6.0/10). Banking and leasing present risk distributed across all document types, while real estate is concentrated on proof of address and payslips. The fraud data guide provides further breakdowns by document type.

How will document fraud in France evolve in 2027-2028?

Projections indicate a continued rise in declared losses (1.72 billion EUR in 2027, 1.85 billion EUR in 2028), an acceleration in the deepfake share of detected forgeries (from 38% to 55%), but also a significant improvement in the detection rate (from 41% to 52%) thanks to increasing adoption of automated verification solutions. Deployment of the European Digital Identity Wallet (eIDAS 2.0) could be a major reduction factor from 2028.

How can businesses reduce their exposure to document fraud in France?

Three immediate levers: (1) Automate document verification with an AI solution capable of analysing metadata, structure and document consistency in real time — reducing detection time from 72 days to under 3 seconds. (2) Prioritise controls by sector and document type using a risk framework like the CheckFile Document Risk Index to concentrate resources on the most critical vectors. (3) Integrate verification into existing workflows (onboarding, subscription, application review) rather than treating it as an ex-post control.

From analysis to action

The 2026 data paints an unambiguous picture: document fraud in France is on a structural upward trajectory, driven by the democratisation of AI tools and the industrialisation of forgery networks. The real cost far exceeds official figures, and the most exposed sectors — banking, leasing, real estate — face risk distributed across the full range of document types.

The CheckFile Document Risk Index provides an actionable framework for prioritising controls and allocating compliance resources where risk is highest. Organisations that integrate automated verification into their processes reduce their detection time from 72 days to under 3 seconds and increase their detection rate from 39% to over 92%.

Discover how CheckFile automatically detects fraudulent documents and calculates your sector risk index. Request a demo

For further reading, see our related analyses:

• Document fraud statistics and trends 2026
• Fraud data guide
• AI document fraud detection techniques
• AMLD6 compliance guide for obliged entities
• Deepfake and synthetic identity documents