Executive Summary

The global identity fraud landscape is undergoing a fundamental transformation, defined by the Sophistication Shift. While the overall volume of fraud attempts has moderated to 2.2% in 2025 from a peak of 2.6% in 2024, this stability masks a dangerous evolution in criminal tactics. Low-effort schemes are being replaced by fewer, sharper, and significantly more damaging attacks, with a 180% year-over-year increase in “sophisticated fraud”. These advanced schemes, now constituting 28% of all fraud attempts compared to just 10% in 2024, leverage coordinated techniques such as synthetic identities, deepfakes, and social engineering.

The primary driver of this shift is the industrialization of fraud by Artificial Intelligence. Generative AI tools (e.g., OpenAI’s Sora 2, Google Veo) are no longer just for simple forgeries; they now power an entire ecosystem that produces near-perfect document replicas, convincing deepfake videos for liveness checks, and scalable synthetic identities. This trend is culminating in the emergence of autonomous AI fraud agents capable of executing entire verification attacks without human intervention.

Fraudsters are also evolving their evasion techniques by attacking the verification process itself through telemetry tampering. By manipulating SDKs, masking device fingerprints with emulator farms, and interfering with camera feeds, attackers are now targeting the context of verification, not just the content.

Key data findings from the report reveal a complex global picture:

  • Top Fraud Types: Identity theft (28%) is the leading third-party fraud, while synthetic identity use (21%) is the top first-party fraud.- Most Targeted Documents: ID cards account for 72% of fraudulent documents by volume, but payment methods now have the highest fraud rate at 6.6%, indicating a strategic pivot toward direct monetization.- Industry Hotspots: Online Media and Dating exhibit the highest fraud rates at 6.3%. The Professional Services sector saw a dramatic 232% year-over-year resurgence in fraud.- Regional Dynamics: Fraud is growing fastest in the Middle East (+19.8% YoY). Africa and the Asia-Pacific region have become dynamic battlegrounds where advanced regulatory efforts in some nations contrast with surging AI-driven fraud in others. The Maldives experienced a staggering 2,100% YoY growth in deepfake attacks, the largest recorded.

The future of fraud prevention requires a paradigm shift from static, document-based checks to continuous, intelligence-driven identity assurance that leverages behavioral analytics, multi-modal AI detection, and cross-channel data intelligence to counter these increasingly sophisticated and automated threats.

1. Introduction and Methodology

The Identity Fraud Report 2025-2026 analyzes the evolution of identity crime, building upon the “Democratization of Fraud” trend identified in the 2024 report. That trend, characterized by the widespread availability of fraud-as-a-service platforms, has matured into the Sophistication Shift, where democratized tools are being repurposed for more targeted, professional, and damaging operations.

This report provides a comprehensive analysis of this shift by combining quantitative data with qualitative insights to prepare businesses, regulators, and consumers for the challenges of 2026 and beyond.

1.1. Methodology

The report’s findings are grounded in a multi-faceted data collection strategy:

  • Internal Data Analysis: The core analysis is based on over 4 million fraud attempts detected on the Sumsub platform, comparing identity verification data from 2024 and 2025. Data from 2023 is included to highlight long-term trends. Analysis is restricted to jurisdictions with over 15,000 verification attempts to ensure statistical reliability.- Fraud Exposure Survey 2025: To deepen the analysis, Sumsub conducted a global survey in August 2025, gathering insights from:300+ fraud and risk professionals from sectors including banking, crypto, payments, e-commerce, and iGaming.- 1,200+ end-users from regions across North America, Latin America, Europe, Asia, Africa, and the Middle East.

All data has been aggregated and anonymized.

2. The Core Theme: The Sophistication Shift

The central trend of 2025 is the Sophistication Shift, a turning point where identity fraud transitions from high-volume, low-effort attacks to fewer, more precise, and highly damaging schemes. While overall fraud rates may appear to stabilize, the underlying threat has become more potent. Every successful attempt now represents greater preparation and results in a more significant impact.

  • Rise in Advanced Attacks: There has been a 180% year-over-year increase in sophisticated fraud, which involves advanced deception techniques, social engineering, and AI-generated identities.- Shifting Composition: In 2024, advanced schemes constituted only 10% of fraud attempts. By 2025, this share has surged to 28%.

This evolution is a direct result of stronger verification platforms filtering out amateurish attempts, forcing fraudsters to adopt more strategic and technologically advanced methods.

“We are witnessing a fundamental shift in the nature of fraud. Generative AI has democratized deception, but it has also forced verification to innovate at a pace faster than ever before. What we’re witnessing now is not a rise in the levels of fraud, but instead, smarter and more deliberate attacks, with multiple layers of deceit.”

— Andrew Sever, CEO at Sumsub

2.1. Defining the Duality of Modern Fraud

The current landscape is characterized by two distinct types of fraudulent activity.

Fraud Type

Description

Common Examples

Simple Fraud

Low-effort, high-volume attempts relying on basic deception. These are often produced using cheap fraud-as-a-service tools and are relatively easy for modern verification systems to detect.

• Poorly edited or stolen ID scans
• Basic document template reuse
• Copy-pasted identity details from data leaks
• Showing a photo to a camera to bypass liveness tests

Sophisticated Fraud

High-effort, coordinated schemes combining multiple advanced techniques. These attacks require planning, technical resources, and often team coordination, making them harder to detect and far more damaging.

• High-fidelity AI-generated IDs paired with deepfake video liveness
• Synthetic identities used to create mule account networks
• Telemetry tampering combined with forged documents
• Orchestrated fraud rings with interacting synthetic and stolen identities

Four dominant trends are accelerating the Sophistication Shift and defining the modern fraud ecosystem.

3.1. Trend 1: AI Industrializes Fraud

In 2025, AI has evolved from a simple tool for forgery into a sophisticated production engine for industrialized, scalable fraud.

  • Advanced Document Forgeries: Image generation tools from platforms like OpenAI can now create IDs with near-perfect replication of fonts, holograms, and textures.- Realistic Synthetic Video: Next-generation text-to-video systems (e.g., Google Veo, OpenAI Sora 2) can render dynamic scenes with realistic facial microexpressions, enabling attackers to stage convincing deepfake liveness checks. The escalating competition between Big Tech companies to release the most realistic AI tools directly fuels the Sophistication Shift.- Automation and Scale: Fraud-as-a-service providers now bundle these AI models into production kits, allowing even low-skilled actors to generate industrial quantities of high-quality forgeries.

3.2. Trend 2: The Emergence of AI Fraud Agents

A new threat vector appeared in 2025: AI fraud agents. These are autonomous systems that combine generative content and behavioral mimicry to execute entire verification attempts from start to finish.

  • Capabilities: These agents can generate fake identities, interact with verification interfaces in real-time, and learn from failed attempts to adapt their strategies.- Future Threat: While still in their infancy, analysts expect a boom in autonomous fraud by 2026, with coordinated fleets of AI agents conducting high-speed, multi-step attacks at scale. This was underscored by a late 2025 report from Anthropic researchers who uncovered a state-linked espionage campaign using autonomous AI agents for cyber operations.

3.3. Trend 3: Telemetry Tampering Becomes the New Evasion

Fraudsters are increasingly targeting the data pipelines of verification systems rather than just the identity artifacts. This “context manipulation” involves manipulating the signals that systems rely on to establish trust.

  • Methods of Attack: Common techniques include SDK and API manipulation, device and environment masking (using emulator farms and virtual machines), and camera feed interference to inject pre-recorded or AI-generated video.- Prevalence of Tools: Developer tools are the most common method for telemetry masking (44% of cases), followed by incognito mode (22%) and privacy-focused browsers (12%).

4. Global Fraud Landscape: Data and Analysis

4.1. Global Fraud Rate Dynamics

The global identity fraud rate has shown volatility but remains at a high level.

  • 2023: 2.0%- 2024: 2.6% (a peak driven by the “Democratization of Fraud”)- 2025: 2.2% (a moderation in volume but an increase in sophistication)

This slight decline should not be mistaken for relief, as the composition of fraud has shifted toward more dangerous, high-quality attacks.

4.2. A Dual Landscape of Fraud Types

In 2025, fraud is best understood by separating it into two categories: first-party fraud (perpetrated by the user themselves) and third-party fraud (perpetrated by external attackers).

First-Party Fraud (Top 5)

Share

Third-Party Fraud (Top 5)

Share

Synthetic Identity Use

21%

Identity Theft

28%

Chargeback Abuse

16%

Account Takeover

19%

Application Fraud

14%

Card Testing

17%

Money Muling

11%

Phishing/Social Engineering

16%

Deepfakes

11%

Bot-driven Attacks

12%

4.3. Analysis of Fraudulent Documents

Forged documents remain a primary entry point for most fraud schemes.

  • **Fraud Share by ID Type:**ID Cards: 72%- Passports: 13%- Driver’s Licenses: 10%- Utility Bills: 2%- Other: 3% Strategic Pivot to Payments: While ID cards dominate in volume, payment methods now have the highest fraud rate of any artifact type at 6.6%. This indicates a strategic shift by fraudsters to move downstream from identity verification to direct financial monetization.Rise of AI-Generated Documents: In 2025, 2% of all detected fake documents were created using generative AI tools like ChatGPT, Grok, and Gemini. This trend, first observed in April 2025, shows a steady upward trajectory, demonstrating how AI is industrializing what was once a niche skill.

5. Industry and Regional Breakdowns

5.1. Identity Fraud by Industry

Fraud pressure varies significantly across different sectors.

Industry

2025 Fraud Rate

YoY Change

Key Insights

Online Media

6.3%

Slight Decrease

Remains a top target for fake account creation and monetization scams.

Dating

6.3%

Dominated by romance scams using AI-generated personas and deepfakes.

Financial Services

2.7%

-2%

Targeted with sophisticated synthetic identities and chargeback abuse.

Crypto

2.2%

+2%

Used as a primary channel for laundering funds from other scams.

Professional Services

1.6%

+232%

A sharp rebound driven by fake credential, invoice, and legal document scams.

Video Gaming

1.6%

-27%

Exposed to account takeover and sophisticated bot-driven item farming.

iGaming

1.2%

+8%

Transformed by deepfakes for age/bonus bypass; synthetics up 329%.

E-commerce

1.0%

-28%

Stronger controls are working, but chargeback and refund fraud persist.

Consumer Trust by Industry

Consumer trust levels correlate with perceived security and regulatory oversight.

  • High Trust: Banking and Financial Services (70/100)- Moderate Trust: Online Shopping (61), Travel Services (60)- Trust Deficit: Crypto (52), Social Media (49), iGaming (49)- Lowest Trust: Dating Platforms (42)

5.2. Regional Breakdowns

Fraud trends are not uniform globally, with significant variations in growth rates and tactics.

Average YoY Fraud Rate Growth by Region (2025):

  • Middle East: +19.8%- APAC: +14.6%- U.S. & Canada: +13.3%- LATAM & Caribbean: +9.3%- Africa: -5.5%- Europe: -16.4%

The Sumsub Global Fraud Index shows that European countries are generally the most protected from fraud, while nations in APAC and Africa are among the least protected.

In-Depth Regional Analysis: Africa

Africa’s expanding digital economy has made it a dynamic battleground where the Sophistication Shift is clearly visible.

  • Key Trends: A surge in selfie-based fraud and deepfake activity. Deepfake attempts grew by +367% YoY in the DRC and +325% in Malawi.- **Country Dynamics:**Rising Hotspots: Mali (+131% YoY) and Côte d’Ivoire (+51% YoY) saw sharp increases tied to mobile money growth outpacing controls.- Declining Markets: Nigeria (-54% YoY) and South Africa (-31% YoY) experienced dramatic drops due to stronger regulation (e.g., NIN-SIM linkage, enhanced AML/CFT frameworks). However, deepfake incidents in South Africa rose by 269%.- Fraud Networks: Zambia has the highest ratio of approved applicants linked to fraud networks (37%). Consumer Insights: Phishing (57%) is the main attack vector. 76.5% of consumers are aware of money muling but underestimate its seriousness.Regulatory Response: Nations like Kenya, Nigeria, and South Africa are actively tightening SIM registration, AML standards, and payment regulations.

In-Depth Regional Analysis: Asia & Pacific (APAC)

APAC is one of the most dynamic fraud environments, reflecting the full spectrum of the Sophistication Shift.

  • Key Trends: A dramatic swing from crude forgeries to AI-driven attacks. Selfie-related fraud now comprises 35.4% of all fraud (+73% YoY), while synthetic data use grew +142% YoY.- **Country Dynamics:**High-Growth Markets: Malaysia saw the highest YoY increase at +197%. Pakistan (5.9% fraud rate) and Sri Lanka (+79% YoY) also saw significant surges.- Declining but Sophisticated Markets: India (-23% YoY), Singapore (-12%), and Hong Kong (-43%) recorded declines due to strong regulation, but the remaining fraud is increasingly advanced. Singapore saw a 158% YoY increase in deepfake incidents. Consumer Insights: Deepfake normalization is high, with 32% having encountered them online. Phishing is the top attack vector (61%).Regulatory Response: Australia has implemented a Scams Prevention Framework Bill and expanded its AML/CTF regime. China has broadened its AML law to cover fraud proceeds.

6. Expert Commentary and Future Outlook

Experts across the industry agree that the nature of fraud and its prevention is at an inflection point.

“By 2026, identity verification will shift from static, document-based checks to continuous, intelligence-driven identity assurance. The biggest trend will be the fusion of behavioral biometrics, device intelligence, and on-chain reputation data, where verification becomes dynamic and adaptive rather than transactional.”

— Dina Mainville, Independent Director at Kraken; Founder & President at Collisionless

Predictions for 2026 and Beyond:

  1. AI-Native Fraud Prevention: Systems will move beyond using AI as a layer to designing platforms where AI autonomously runs detection, decisioning, and learning loops.2. Continuous Identity Assurance: The “check once, trust forever” model will be replaced by continuous monitoring of behavior and transactions throughout the customer lifecycle.3. Rise of Digital Identity Ecosystems: Private and public identity systems will merge, enabling seamless, high-assurance authentication.4. Focus on Regional Intelligence: As fraud tactics become more localized, region-specific verification intelligence will be critical, especially in emerging markets like Africa and Southeast Asia.5. Regulatory Scrutiny: Regulators will demand measurable fraud-loss prevention outcomes over simple box-ticking compliance. The EU AI Act and PSD3/PSR framework are key examples of this tightening oversight.