Document Fraud Statistics and Trends 2026: UK Data and Global Insights
Document fraud costs UK businesses GBP 1.8 billion in 2026. Latest statistics from UK Finance, Cifas, and the NCA. Deepfake trends, sector breakdown, and 5-year evolution data.
Summarize this article with
ChatGPT
Claude
Perplexity
Gemini
GrokDocument fraud cost UK businesses an estimated GBP 1.8 billion in 2025, according to data compiled from UK Finance, Cifas, and the National Crime Agency. That figure captures only reported losses. The true cost, including undetected fraud, is estimated at three times that amount. This article compiles the latest country-specific data, tracks the five-year trajectory of document fraud in the UK, and analyses the accelerating role of AI-generated forgeries.
For foundational statistics on document fraud, see our fraud data guide. For a deep dive into detection methods, read our analysis of AI document fraud detection.
GBP 1.8 billion: the annual cost of document fraud to UK businesses
UK businesses lost an estimated GBP 1.8 billion to document fraud in 2025, up 17% from GBP 1.54 billion in 2023. This estimate draws on the UK Finance Annual Fraud Report 2025, which recorded GBP 1.17 billion in authorised and unauthorised payment fraud alone, with document-facilitated fraud representing a substantial share.
Action Fraud received over 374,000 fraud reports in the year ending March 2025, a 16% increase year-on-year. While not all reports involve document fraud specifically, Cifas data indicates that 43% of all fraud cases reported to their National Fraud Database involved some form of document falsification.
What changed in 2025-2026
Three structural shifts are driving the increase:
- AI-generated documents: 34% of forged documents flagged by UK financial institutions in 2025 showed indicators of AI generation, up from 9% in 2023 (Cifas Fraudscape 2025).
- Organised fraud networks: The NCA disrupted 23 document fraud networks in 2025, but estimates suggest at least 40 active networks target the UK market.
- Digital-first processes: 82% of customer onboarding in UK financial services is now fully digital, creating new attack surfaces for document fraud.
Table 1: Document fraud types by volume and cost in the UK
| Fraud type | Share of detected cases (2025) | Average cost per incident | Primary sectors |
|---|---|---|---|
| Forged proof of address | 21% | GBP 7,200 | Banking, letting, insurance |
| Fake payslips and P60s | 19% | GBP 11,400 | Mortgage, rental, consumer credit |
| Manipulated financial statements | 14% | GBP 74,000 | Commercial lending, leasing |
| Forged Companies House documents | 12% | GBP 29,000 | B2B, trade credit, procurement |
| Identity fraud via fake IDs/passports | 13% | GBP 16,800 | Banking, telecoms, gambling |
| Fraudulent certificates (insurance, HMRC) | 11% | GBP 19,500 | Construction, subcontracting |
| Manipulated bank details | 10% | GBP 42,000 | All sectors (APP fraud) |
Sources: UK Finance Annual Fraud Report 2025, Cifas Fraudscape 2025, HMRC fraud statistics.
Forged proof of address and fake payslips together account for 40% of all detected cases. However, manipulated financial statements and bank details generate the highest per-incident losses, reflecting their use in high-value commercial fraud.
374,000 fraud reports filed with Action Fraud in 2025
Action Fraud data shows a steady increase in fraud reporting. Of the 374,000 reports filed in the year ending March 2025, the National Fraud Intelligence Bureau (NFIB) identified document fraud as a contributing factor in approximately 161,000 cases.
By sector exposure
UK financial services bears the heaviest burden. UK Finance reports that authorised push payment (APP) fraud alone reached GBP 459 million in 2024, with many cases originating from fraudulent supporting documents. The FCA issued GBP 124 million in fines to firms with inadequate fraud controls in 2025.
Table 2: UK document fraud evolution (2021-2026)
| Indicator | 2021 | 2022 | 2023 | 2024 | 2025 | 2026 (est.) |
|---|---|---|---|---|---|---|
| Estimated annual cost (GBP bn) | 1.21 | 1.38 | 1.54 | 1.67 | 1.80 | 1.92 |
| Cifas fraud cases involving documents | 38% | 39% | 41% | 42% | 43% | 45% |
| Detection rate (financial services) | 31% | 34% | 36% | 38% | 41% | 43% |
| AI-generated forgeries (share of detected) | < 1% | 3% | 9% | 18% | 34% | 42% |
| Action Fraud reports (thousands) | 298 | 312 | 322 | 347 | 374 | 398 |
| Average detection time (days) | 104 | 96 | 89 | 82 | 74 | 68 |
Sources: UK Finance, Cifas, Action Fraud, ACFE Report to the Nations 2024.
The five-year trend shows a consistent pattern: fraud volumes grow 8-12% per year, but detection rates improve incrementally. The most striking shift is in AI-generated forgeries, which have grown from less than 1% of detected fakes in 2021 to an estimated 42% in 2026.
AI-generated forgeries: 42% of detected fakes in 2026
The Cifas Fraudscape 2025 report flagged AI-generated document fraud as the single fastest-growing fraud vector. 34% of forged documents flagged by member organisations in 2025 exhibited characteristics of AI generation, including synthetic metadata patterns, uniform font rendering inconsistent with scanned originals, and missing native editing layers.
How AI forgeries work in practice
- Full generation: A document created entirely by a generative model, replicating the layout, fonts, and formatting of a genuine original. Common for utility bills, payslips, and insurance certificates.
- Targeted modification: An authentic document with specific fields (amounts, dates, names) altered using AI tools. Harder to detect because the overall structure remains genuine.
- Template cloning: Reproduction of letterheads, stamps, or watermarks from scanned originals. Used on Companies House certificates and HMRC correspondence.
The NCA's 2025 threat assessment specifically identified generative AI as an "enabler of document fraud at scale," noting that the cost of producing a convincing forged document has fallen by approximately 85% since 2021.
Most exposed sectors in the UK
Financial services: 36% of detected document fraud
UK banks and building societies detected 36% of all document fraud cases reported to Cifas in 2025. The FCA has made document verification a priority enforcement area, issuing 14 warning notices related to inadequate KYC document controls. Compliance fines for document verification failures increased by 41% year-on-year.
Lettings and real estate: 22% of detected document fraud
Fake payslips and forged references are endemic in the UK rental market. Industry data suggests 1 in 6 rental applications in London contains at least one manipulated document. The Tenant Fees Act and right-to-rent checks have increased the volume of documents required, creating more opportunities for fraud.
Commercial lending and leasing: 16% of detected document fraud
Manipulated financial statements and forged Companies House certificates drive the highest per-incident losses. A single falsified set of accounts can lead to credit facilities of GBP 500,000 or more being extended to insolvent businesses.
Acceleration factors: why document fraud is growing
1. Generative AI lowers the barrier to entry
The cost and skill required to produce a convincing forged document have collapsed. Tools that generate realistic payslips, utility bills, and bank statements are available for less than GBP 50 on dark web marketplaces. The NCA estimates that 60% of document fraud now involves some degree of AI assistance.
2. Digital onboarding without adequate controls
82% of customer onboarding in UK financial services is now digital, but only 43% of firms use automated document verification. The gap between digital processes and verification capabilities creates systematic vulnerability.
3. Cross-border fraud networks
The NCA identified 23 organised networks producing forged documents targeting UK businesses in 2025. These networks operate from multiple jurisdictions and offer "complete identity packages" (ID, proof of address, income documents) for GBP 400-2,500.
Regulatory response: enforcement is tightening
The UK regulatory environment is responding to the document fraud surge:
- Economic Crime and Corporate Transparency Act 2023: strengthened Companies House verification requirements and created new offences for false filing.
- FCA Consumer Duty: firms must demonstrate reasonable steps to verify customer documents.
- UK Finance Confirmation of Payee: mandatory for all payment service providers, reducing but not eliminating APP fraud enabled by forged documents.
The total value of regulatory fines for compliance failures reached GBP 189 million in 2025.
FAQ
How much does document fraud cost UK businesses in 2026?
The estimated annual cost is GBP 1.8 billion in detected losses (2025 data), projected to reach GBP 1.92 billion in 2026. Including undetected fraud, the true cost likely exceeds GBP 5 billion.
What percentage of UK fraud involves forged documents?
43% of all fraud cases reported to the Cifas National Fraud Database in 2025 involved some form of document falsification. In financial services specifically, the proportion reaches 52%.
How fast are AI-generated forgeries growing?
AI-generated forgeries represented 34% of detected fakes in 2025, up from less than 1% in 2021. The 2026 projection is 42%. This represents the fastest-growing fraud vector in the UK.
Which UK sectors are most affected by document fraud?
Financial services accounts for 36% of detected document fraud, followed by lettings and real estate (22%) and commercial lending (16%). Financial services also faces the heaviest regulatory consequences for detection failures.
What is the average time to detect document fraud?
The average detection time has fallen from 104 days in 2021 to 74 days in 2025. Organisations using automated AI verification reduce this to under 5 seconds at the point of document submission.
Conclusion: UK document fraud outpaces manual detection
The 2026 data confirms a structural trend: document fraud volumes grow faster than manual detection capabilities. The organisations narrowing the gap are those deploying automated verification at the point of document intake.
For further reading, explore our fraud data guide and our breakdown of AI-powered fraud detection techniques.